[upd] Implement BDI reasoning system and enhance AI decision-making with persistent beliefs and desires. Introduced Redis state storage for improved simulation state management and added asynchronous logging for performance optimization. Updated configuration parameters to support new features.

README.md

@@ -9,14 +9,14 @@ This project simulates a village economy with autonomous AI agents. Each agent h
 ### Features
 
 - **Agent-based simulation**: Multiple AI agents with different professions
+- **GOAP AI system**: Goal-Oriented Action Planning for intelligent agent behavior
 - **Vital stats system**: Energy, Hunger, Thirst, and Heat with passive decay
 - **Market economy**: Order book system for trading resources
 - **Day/Night cycle**: 10 day steps + 1 night step per day
-- **Maslow-priority AI**: Agents prioritize survival over economic activities
-- **Real-time visualization**: Pygame frontend showing agents and their states
+- **Real-time visualization**: Web-based frontend showing agents and their states
 - **Agent movement**: Agents visually move to different locations based on their actions
 - **Action indicators**: Visual feedback showing what each agent is doing
-- **Settings panel**: Adjust simulation parameters with sliders
+- **GOAP Debug Panel**: View agent planning and decision-making in real-time
 - **Detailed logging**: All simulation steps are logged for analysis
 
 ## Architecture
@@ -28,11 +28,13 @@ villsim/
 │   ├── config.py           # Centralized configuration
 │   ├── api/                # REST API endpoints
 │   ├── core/               # Game logic (engine, world, market, AI, logger)
+│   │   └── goap/           # GOAP AI system (planner, actions, goals)
 │   └── domain/             # Data models (agent, resources, actions)
-├── frontend/               # Pygame visualizer
-│   ├── main.py             # Entry point
-│   ├── client.py           # HTTP client
-│   └── renderer/           # Drawing components (map, agents, UI, settings)
+├── web_frontend/           # Web-based visualizer
+│   ├── index.html          # Main application
+│   ├── goap_debug.html     # GOAP debugging view
+│   └── src/                # JavaScript modules (scenes, API client)
+├── tools/                  # Analysis and optimization scripts
 ├── logs/                   # Simulation log files (created on run)
 ├── docs/design/            # Design documents
 ├── requirements.txt
@@ -79,40 +81,25 @@ The server will start at `http://localhost:8000`. You can access:
 - API docs: `http://localhost:8000/docs`
 - Health check: `http://localhost:8000/health`
 
-### Start the Frontend Visualizer
+### Start the Web Frontend
 
-Open another terminal and run:
+Open the web frontend by opening `web_frontend/index.html` in a web browser, or serve it with a local HTTP server:
 
 ```bash
-python -m frontend.main
+cd web_frontend
+python -m http.server 8080
 ```
 
-A Pygame window will open showing the simulation.
+Then navigate to `http://localhost:8080` in your browser.
 
 ## Controls
 
-| Key | Action |
-|-----|--------|
-| `SPACE` | Advance one turn (manual mode) |
-| `R` | Reset simulation |
-| `M` | Toggle between MANUAL and AUTO mode |
-| `S` | Open/close settings panel |
-| `ESC` | Close settings or quit |
+The web frontend provides buttons for:
+- **Step**: Advance one turn (manual mode)
+- **Auto/Manual**: Toggle between automatic and manual mode
+- **Reset**: Reset simulation
 
-Hover over agents to see detailed information.
-
-## Settings Panel
-
-Press `S` to open the settings panel where you can adjust:
-
-- **Agent Stats**: Max values and decay rates for energy, hunger, thirst, heat
-- **World Settings**: Grid size, initial agent count, day length
-- **Action Costs**: Energy costs for hunting, gathering, etc.
-- **Resource Effects**: How much stats are restored by consuming resources
-- **Market Settings**: Price adjustment timing and rates
-- **Simulation Speed**: Auto-step interval
-
-Changes require clicking "Apply & Restart" to take effect.
+Click on agents to see detailed information. Use the GOAP debug panel (`goap_debug.html`) to inspect agent planning.
 
 ## Logging
 
@@ -189,12 +176,14 @@ Action indicators above agents show:
 - Movement animation when traveling
 - Dotted line to destination
 
-### AI Priority System
+### AI System (GOAP)
 
-1. **Critical needs** (stat < 20%): Consume, buy, or gather resources
-2. **Energy management**: Rest if too tired
-3. **Economic activity**: Sell excess inventory, buy needed materials
-4. **Routine work**: Perform profession-specific tasks
+The simulation uses Goal-Oriented Action Planning (GOAP) for intelligent agent behavior:
+
+1. **Goals**: Agents have weighted goals (Survive, Maintain Heat, Build Wealth, etc.)
+2. **Actions**: Agents can perform actions with preconditions and effects
+3. **Planning**: A* search finds optimal action sequences to satisfy goals
+4. **Personality**: Each agent has unique traits affecting goal weights and decisions
 
 ## Development
 
@@ -202,9 +191,10 @@ Action indicators above agents show:
 
 - **Config** (`backend/config.py`): Centralized configuration with dataclasses
 - **Domain Layer** (`backend/domain/`): Pure data models
-- **Core Layer** (`backend/core/`): Game logic, AI, market, logging
+- **Core Layer** (`backend/core/`): Game logic, market, logging
+- **GOAP AI** (`backend/core/goap/`): Goal-oriented action planning system
 - **API Layer** (`backend/api/`): FastAPI routes and schemas
-- **Frontend** (`frontend/`): Pygame visualization client
+- **Web Frontend** (`web_frontend/`): Browser-based visualization
 
 ### Analyzing Logs
 
@@ -227,7 +217,7 @@ with open("logs/sim_20260118_123456.jsonl") as f:
 - Agent reproduction
 - Skill progression
 - Persistent save/load
-- Web-based frontend alternative
+- Unity frontend integration
 
 ## License
 

backend/api/routes.py

@@ -93,12 +93,12 @@ def get_market_prices():
 )
 def initialize_simulation(request: InitializeRequest = None):
     """Initialize or reset the simulation.
-    
+
     If request is provided with specific values, use those.
     Otherwise, use values from config.json.
     """
     engine = get_engine()
-    
+
     if request and (request.num_agents != 8 or request.world_width != 20 or request.world_height != 20):
         # Custom values provided - use them
         from backend.core.world import WorldConfig
@@ -113,7 +113,7 @@ def initialize_simulation(request: InitializeRequest = None):
         # Use values from config.json
         engine.reset()  # This now loads from config.json automatically
         num_agents = engine.world.config.initial_agents
-    
+
     return ControlResponse(
         success=True,
         message=f"Simulation initialized with {num_agents} agents",
@@ -131,21 +131,21 @@ def initialize_simulation(request: InitializeRequest = None):
 def next_step():
     """Advance the simulation by one step."""
     engine = get_engine()
-    
+
     if engine.mode == SimulationMode.AUTO:
         raise HTTPException(
             status_code=400,
             detail="Cannot manually advance while in AUTO mode. Switch to MANUAL first.",
         )
-    
+
     if not engine.is_running:
         raise HTTPException(
             status_code=400,
             detail="Simulation is not running. Initialize first.",
         )
-    
+
     turn_log = engine.next_step()
-    
+
     return ControlResponse(
         success=True,
         message=f"Advanced to turn {engine.world.current_turn}",
@@ -163,7 +163,7 @@ def next_step():
 def set_mode(request: SetModeRequest):
     """Set the simulation mode."""
     engine = get_engine()
-    
+
     try:
         mode = SimulationMode(request.mode)
         engine.set_mode(mode)
@@ -223,11 +223,11 @@ def get_logs(limit: int = 10):
 def get_turn_log(turn: int):
     """Get log for a specific turn."""
     engine = get_engine()
-    
+
     for log in engine.turn_logs:
         if log.turn == turn:
             return log.to_dict()
-    
+
     raise HTTPException(status_code=404, detail=f"Log for turn {turn} not found")
 
 
@@ -315,3 +315,107 @@ def load_config_from_file():
     except Exception as e:
         raise HTTPException(status_code=500, detail=f"Failed to load config: {str(e)}")
 
+
+# ============== GOAP Debug Endpoints ==============
+
+@router.get(
+    "/goap/debug/{agent_id}",
+    summary="Get GOAP debug info for an agent",
+    description="Returns detailed GOAP decision-making info including goals, actions, and plans.",
+)
+def get_agent_goap_debug(agent_id: str):
+    """Get GOAP debug information for a specific agent."""
+    engine = get_engine()
+    agent = engine.world.get_agent(agent_id)
+
+    if agent is None:
+        raise HTTPException(status_code=404, detail=f"Agent {agent_id} not found")
+
+    if not agent.is_alive():
+        raise HTTPException(status_code=400, detail=f"Agent {agent_id} is not alive")
+
+    from backend.core.goap.debug import get_goap_debug_info
+
+    debug_info = get_goap_debug_info(
+        agent=agent,
+        market=engine.market,
+        step_in_day=engine.world.step_in_day,
+        day_steps=engine.world.config.day_steps,
+        is_night=engine.world.is_night(),
+    )
+
+    return debug_info.to_dict()
+
+
+@router.get(
+    "/goap/debug",
+    summary="Get GOAP debug info for all agents",
+    description="Returns GOAP decision-making info for all living agents.",
+)
+def get_all_goap_debug():
+    """Get GOAP debug information for all living agents."""
+    engine = get_engine()
+
+    from backend.core.goap.debug import get_all_agents_goap_debug
+
+    debug_infos = get_all_agents_goap_debug(
+        agents=engine.world.agents,
+        market=engine.market,
+        step_in_day=engine.world.step_in_day,
+        day_steps=engine.world.config.day_steps,
+        is_night=engine.world.is_night(),
+    )
+
+    return {
+        "agents": [info.to_dict() for info in debug_infos],
+        "count": len(debug_infos),
+        "current_turn": engine.world.current_turn,
+        "is_night": engine.world.is_night(),
+    }
+
+
+@router.get(
+    "/goap/goals",
+    summary="Get all GOAP goals",
+    description="Returns a list of all available GOAP goals.",
+)
+def get_goap_goals():
+    """Get all available GOAP goals."""
+    from backend.core.goap import get_all_goals
+
+    goals = get_all_goals()
+    return {
+        "goals": [
+            {
+                "name": g.name,
+                "type": g.goal_type.value,
+                "max_plan_depth": g.max_plan_depth,
+            }
+            for g in goals
+        ],
+        "count": len(goals),
+    }
+
+
+@router.get(
+    "/goap/actions",
+    summary="Get all GOAP actions",
+    description="Returns a list of all available GOAP actions.",
+)
+def get_goap_actions():
+    """Get all available GOAP actions."""
+    from backend.core.goap import get_all_actions
+
+    actions = get_all_actions()
+    return {
+        "actions": [
+            {
+                "name": a.name,
+                "action_type": a.action_type.value,
+                "target_resource": a.target_resource.value if a.target_resource else None,
+            }
+            for a in actions
+        ],
+        "count": len(actions),
+    }
+

backend/api/schemas.py

@@ -55,9 +55,31 @@ class AgentResponse(BaseModel):
     inventory: list[ResourceSchema]
     money: int
     is_alive: bool
+    is_corpse: bool = False
     can_act: bool
     current_action: AgentActionSchema
     last_action_result: str
+    death_turn: int = -1
+    death_reason: str = ""
+    # Age system
+    age: int = 25
+    max_age: int = 70
+    age_category: str = "prime"
+    birth_day: int = 0
+    generation: int = 0
+    parent_ids: list[str] = []
+    children_count: int = 0
+    # Age modifiers
+    skill_modifier: float = 1.0
+    energy_cost_modifier: float = 1.0
+    learning_modifier: float = 1.0
+    # Personality and skills
+    personality: dict = {}
+    skills: dict = {}
+    actions_performed: dict = {}
+    total_trades: int = 0
+    total_money_earned: int = 0
+    action_history: list = []
 
 
 # ============== Market Schemas ==============
@@ -119,6 +141,7 @@ class StatisticsSchema(BaseModel):
     living_agents: int
     total_agents_spawned: int
     total_agents_died: int
+    total_births: int = 0
     total_money_in_circulation: int
     professions: dict[str, int]
     # Wealth inequality metrics
@@ -127,6 +150,16 @@ class StatisticsSchema(BaseModel):
     richest_agent: int = 0
     poorest_agent: int = 0
     gini_coefficient: float = 0.0
+    # Age demographics
+    age_distribution: dict[str, int] = {}
+    avg_age: float = 0.0
+    oldest_agent: int = 0
+    youngest_agent: int = 0
+    generations: dict[int, int] = {}
+    # Death statistics
+    deaths_by_cause: dict[str, int] = {}
+    # Village storage
+    village_storage: dict[str, int] = {}
 
 
 class ActionLogSchema(BaseModel):
@@ -142,10 +175,12 @@ class TurnLogSchema(BaseModel):
     turn: int
     agent_actions: list[ActionLogSchema]
     deaths: list[str]
+    births: list[str] = []
     trades: list[dict]
     resources_produced: dict[str, int] = {}
     resources_consumed: dict[str, int] = {}
     resources_spoiled: dict[str, int] = {}
+    day_events: dict = {}
 
 
 class ResourceStatsSchema(BaseModel):

backend/config.py

@@ -14,19 +14,19 @@ class AgentStatsConfig:
     max_hunger: int = 100
     max_thirst: int = 100  # Increased from 50 to give more buffer
     max_heat: int = 100
-    
+
     # Starting values
     start_energy: int = 50
     start_hunger: int = 80
     start_thirst: int = 80  # Increased from 40 to start with more buffer
     start_heat: int = 100
-    
+
     # Decay rates per turn
     energy_decay: int = 2
     hunger_decay: int = 2
     thirst_decay: int = 2  # Reduced from 3 to match hunger decay rate
     heat_decay: int = 2
-    
+
     # Thresholds
     critical_threshold: float = 0.25  # 25% triggers survival mode
     low_energy_threshold: int = 15    # Minimum energy to work
@@ -39,7 +39,7 @@ class ResourceConfig:
     meat_decay: int = 8   # Increased from 5 to give more time to use
     berries_decay: int = 25
     clothes_decay: int = 50
-    
+
     # Resource effects
     meat_hunger: int = 30
     meat_energy: int = 5
@@ -62,11 +62,11 @@ class ActionConfig:
     weave_energy: int = -8
     build_fire_energy: int = -5
     trade_energy: int = -1
-    
+
     # Success chances (0.0 to 1.0)
     hunt_success: float = 0.7
     chop_wood_success: float = 0.9
-    
+
     # Output quantities
     hunt_meat_min: int = 1
     hunt_meat_max: int = 3
@@ -86,7 +86,7 @@ class WorldConfig:
     initial_agents: int = 8
     day_steps: int = 10
     night_steps: int = 1
-    
+
     # Agent configuration
     inventory_slots: int = 10
     starting_money: int = 100
@@ -103,74 +103,287 @@ class MarketConfig:
 @dataclass
 class EconomyConfig:
     """Configuration for economic behavior and agent trading.
-    
+
     These values control how agents perceive the value of money and trading.
     Higher values make agents more trade-oriented.
     """
     # How much agents value money vs energy
     # Higher = agents see money as more valuable (trade more)
-    energy_to_money_ratio: float = 1.5  # 1 energy ≈ 1.5 coins
-    
+    energy_to_money_ratio: float = 150  # 1 energy ≈ 150 coins
+
+    # Minimum price floor for any market transaction
+    min_price: int = 100
+
     # How strongly agents desire wealth (0-1)
     # Higher = agents will prioritize building wealth
     wealth_desire: float = 0.3
-    
+
     # Buy efficiency threshold (0-1)
     # If market price < (threshold * fair_value), buy instead of gather
     # 0.7 means: buy if price is 70% or less of the fair value
     buy_efficiency_threshold: float = 0.7
-    
+
     # Minimum wealth target - agents want at least this much money
-    min_wealth_target: int = 50
-    
+    min_wealth_target: int = 5000
+
     # Price adjustment limits
     max_price_markup: float = 2.0      # Maximum price = 2x base value
     min_price_discount: float = 0.5    # Minimum price = 50% of base value
 
 
+@dataclass
+class AIConfig:
+    """Configuration for AI decision-making system."""
+    # Maximum A* iterations for GOAP planner
+    goap_max_iterations: int = 50
+
+    # Maximum plan depth (number of actions in a plan)
+    goap_max_plan_depth: int = 3
+
+    # Fall back to reactive planning if GOAP fails to find a plan
+    reactive_fallback: bool = True
+
+    # Use BDI (Belief-Desire-Intention) instead of pure GOAP
+    # BDI adds persistent beliefs, long-term desires, and plan commitment
+    use_bdi: bool = False
+
+
+@dataclass
+class BDIConfig:
+    """Configuration for BDI (Belief-Desire-Intention) reasoning system.
+    
+    BDI extends GOAP with:
+    - Persistent beliefs (memory of past events)
+    - Long-term desires (personality-driven motivations)
+    - Committed intentions (plan persistence)
+    """
+    # Timeslicing: how often agents run full deliberation
+    # 1 = every turn, 3 = every 3rd turn (staggered by agent ID)
+    thinking_interval: int = 1
+
+    # Maximum consecutive action failures before replanning
+    max_consecutive_failures: int = 2
+
+    # Priority multiplier needed to switch from current intention
+    # 1.5 = new goal must be 50% higher priority to cause a switch
+    priority_switch_threshold: float = 1.5
+
+    # Memory system settings
+    memory_max_events: int = 50      # Max events to remember
+    memory_decay_rate: float = 0.1   # How fast memories fade
+
+
+@dataclass
+class RedisConfig:
+    """Configuration for optional Redis state storage.
+    
+    Redis enables:
+    - Persistent state across restarts
+    - Decoupled UI polling (web clients read independently)
+    - Distributed access (multiple simulation instances)
+    """
+    enabled: bool = False
+    host: str = "localhost"
+    port: int = 6379
+    db: int = 0
+    password: Optional[str] = None
+    prefix: str = "villsim:"
+    ttl_seconds: int = 3600  # 1 hour default TTL
+
+
+@dataclass
+class AgeConfig:
+    """Configuration for the age and lifecycle system.
+
+    Age affects skills, energy costs, and creates birth/death cycles.
+    Age is measured in "years" where 1 year = 1 simulation day.
+
+    Population is controlled by economy:
+    - Birth rate scales with village prosperity (food availability)
+    - Parents transfer wealth to children at birth and death
+    """
+    # Starting age range for initial agents
+    min_start_age: int = 18
+    max_start_age: int = 35
+
+    # Age category thresholds
+    young_age_threshold: int = 25      # Below this = young
+    prime_age_start: int = 25          # Prime age begins
+    prime_age_end: int = 50            # Prime age ends
+    old_age_threshold: int = 50        # Above this = old
+
+    # Lifespan
+    base_max_age: int = 75             # Base maximum age
+    max_age_variance: int = 10         # ± variance for max age
+    age_per_day: int = 1               # How many "years" per sim day
+
+    # Birth mechanics - economy controlled
+    birth_cooldown_days: int = 20      # Days after birth before can birth again
+    min_birth_age: int = 20            # Minimum age to give birth
+    max_birth_age: int = 45            # Maximum age to give birth
+    birth_base_chance: float = 0.02    # Base chance of birth per day
+    birth_prosperity_multiplier: float = 3.0  # Max multiplier based on food abundance
+    birth_food_requirement: int = 60   # Min hunger to attempt birth
+    birth_energy_requirement: int = 25 # Min energy to attempt birth
+
+    # Wealth transfer
+    birth_wealth_transfer: float = 0.25  # Parent gives 25% wealth to child at birth
+    inheritance_enabled: bool = True     # Children inherit from dead parents
+    child_start_age: int = 18           # Age children start at (adult)
+
+    # Age modifiers for YOUNG agents (learning phase)
+    young_skill_multiplier: float = 0.8    # Skills are 80% effective
+    young_learning_multiplier: float = 1.4 # Learn 40% faster
+    young_energy_cost_multiplier: float = 0.85  # 15% less energy cost
+
+    # Age modifiers for PRIME agents (peak performance)
+    prime_skill_multiplier: float = 1.0
+    prime_learning_multiplier: float = 1.0
+    prime_energy_cost_multiplier: float = 1.0
+
+    # Age modifiers for OLD agents (wisdom but frailty)
+    old_skill_multiplier: float = 1.15     # Skills 15% more effective (wisdom)
+    old_learning_multiplier: float = 0.6   # Learn 40% slower
+    old_energy_cost_multiplier: float = 1.2   # 20% more energy cost
+    old_max_energy_multiplier: float = 0.75   # 25% less max energy
+    old_decay_multiplier: float = 1.15        # 15% faster stat decay
+
+
+@dataclass
+class StorageConfig:
+    """Configuration for resource storage limits.
+
+    Limits the total resources that can exist in the village economy.
+    """
+    # Village-wide storage limits per resource type
+    village_meat_limit: int = 100
+    village_berries_limit: int = 150
+    village_water_limit: int = 200
+    village_wood_limit: int = 200
+    village_hide_limit: int = 80
+    village_clothes_limit: int = 50
+
+    # Market limits
+    market_order_limit_per_agent: int = 5   # Max active orders per agent
+    market_total_order_limit: int = 500     # Max total market orders
+
+
+@dataclass
+class SinksConfig:
+    """Configuration for resource sinks (ways resources leave the economy).
+
+    These create pressure to keep producing resources rather than hoarding.
+    """
+    # Daily decay of village storage (percentage)
+    daily_village_decay_rate: float = 0.02  # 2% of stored resources decay daily
+
+    # Money tax (redistributed or removed)
+    daily_tax_rate: float = 0.01            # 1% wealth tax per day
+
+    # Random events
+    random_event_chance: float = 0.05       # 5% chance of event per day
+    fire_event_resource_loss: float = 0.1   # 10% resources lost in fire
+    theft_event_money_loss: float = 0.05    # 5% money stolen
+
+    # Maintenance costs
+    clothes_maintenance_per_day: int = 1    # Clothes degrade 1 durability/day
+    fire_wood_cost_per_night: int = 1       # Wood consumed to stay warm at night
+
+
+@dataclass
+class PerformanceConfig:
+    """Configuration for performance optimization.
+
+    Controls logging and memory usage to keep simulation fast at high turn counts.
+    """
+    # Logging control
+    logging_enabled: bool = False           # Enable file logging (disable for speed)
+    detailed_logging: bool = False          # Enable verbose per-agent logging
+    async_logging: bool = True              # Use non-blocking background logging
+    log_flush_interval: int = 50            # Flush logs every N turns (not every turn)
+
+    # Memory management
+    max_turn_logs: int = 100                # Keep only last N turn logs in memory
+
+    # Statistics calculation frequency
+    stats_update_interval: int = 10         # Update expensive stats every N turns
+
+    # State storage
+    state_storage_enabled: bool = True      # Enable state snapshotting
+
+
 @dataclass
 class SimulationConfig:
     """Master configuration containing all sub-configs."""
+    performance: PerformanceConfig = field(default_factory=PerformanceConfig)
     agent_stats: AgentStatsConfig = field(default_factory=AgentStatsConfig)
     resources: ResourceConfig = field(default_factory=ResourceConfig)
     actions: ActionConfig = field(default_factory=ActionConfig)
     world: WorldConfig = field(default_factory=WorldConfig)
     market: MarketConfig = field(default_factory=MarketConfig)
     economy: EconomyConfig = field(default_factory=EconomyConfig)
-    
+    ai: AIConfig = field(default_factory=AIConfig)
+    bdi: BDIConfig = field(default_factory=BDIConfig)
+    redis: RedisConfig = field(default_factory=RedisConfig)
+    age: AgeConfig = field(default_factory=AgeConfig)
+    storage: StorageConfig = field(default_factory=StorageConfig)
+    sinks: SinksConfig = field(default_factory=SinksConfig)
+
     # Simulation control
     auto_step_interval: float = 1.0  # Seconds between auto steps
-    
+
     def to_dict(self) -> dict:
         """Convert to dictionary."""
-        return {
+        result = {
+            "performance": asdict(self.performance),
+            "ai": asdict(self.ai),
+            "bdi": asdict(self.bdi),
             "agent_stats": asdict(self.agent_stats),
             "resources": asdict(self.resources),
             "actions": asdict(self.actions),
             "world": asdict(self.world),
             "market": asdict(self.market),
             "economy": asdict(self.economy),
+            "age": asdict(self.age),
+            "storage": asdict(self.storage),
+            "sinks": asdict(self.sinks),
             "auto_step_interval": self.auto_step_interval,
         }
-    
+        # Handle redis separately due to Optional field
+        redis_dict = asdict(self.redis)
+        result["redis"] = redis_dict
+        return result
+
     @classmethod
     def from_dict(cls, data: dict) -> "SimulationConfig":
         """Create from dictionary."""
+        # Handle redis config specially due to Optional password
+        redis_data = data.get("redis", {})
+        if redis_data.get("password") is None:
+            redis_data["password"] = None
+
         return cls(
+            performance=PerformanceConfig(**data.get("performance", {})),
+            ai=AIConfig(**data.get("ai", {})),
+            bdi=BDIConfig(**data.get("bdi", {})),
+            redis=RedisConfig(**redis_data),
             agent_stats=AgentStatsConfig(**data.get("agent_stats", {})),
             resources=ResourceConfig(**data.get("resources", {})),
             actions=ActionConfig(**data.get("actions", {})),
             world=WorldConfig(**data.get("world", {})),
             market=MarketConfig(**data.get("market", {})),
             economy=EconomyConfig(**data.get("economy", {})),
+            age=AgeConfig(**data.get("age", {})),
+            storage=StorageConfig(**data.get("storage", {})),
+            sinks=SinksConfig(**data.get("sinks", {})),
             auto_step_interval=data.get("auto_step_interval", 1.0),
         )
-    
+
     def save(self, path: str = "config.json") -> None:
         """Save configuration to JSON file."""
         with open(path, "w") as f:
             json.dump(self.to_dict(), f, indent=2)
-    
+
     @classmethod
     def load(cls, path: str = "config.json") -> "SimulationConfig":
         """Load configuration from JSON file."""
@@ -188,7 +401,7 @@ _config: Optional[SimulationConfig] = None
 
 def get_config() -> SimulationConfig:
     """Get the global configuration instance.
-    
+
     Loads from config.json if not already loaded.
     """
     global _config
@@ -246,10 +459,21 @@ def _reset_all_caches() -> None:
         reset_action_config_cache()
     except ImportError:
         pass
-    
+
     try:
         from backend.domain.resources import reset_resource_cache
         reset_resource_cache()
     except ImportError:
         pass
 
+    try:
+        from backend.core.ai import reset_ai_config_cache
+        reset_ai_config_cache()
+    except ImportError:
+        pass
+
+    try:
+        from backend.core.storage import reset_state_store
+        reset_state_store()
+    except ImportError:
+        pass

backend/core/__init__.py

@@ -3,7 +3,6 @@
 from .world import World, TimeOfDay
 from .market import Order, OrderBook
 from .engine import GameEngine, SimulationMode
-from .ai import AgentAI
 from .logger import SimulationLogger, get_simulation_logger
 
 __all__ = [
@@ -13,7 +12,6 @@ __all__ = [
     "OrderBook",
     "GameEngine",
     "SimulationMode",
-    "AgentAI",
     "SimulationLogger",
     "get_simulation_logger",
 ]

backend/core/bdi/__init__.py

@@ -0,0 +1,50 @@
+"""BDI (Belief-Desire-Intention) module for agent AI.
+
+This module provides a BDI architecture that wraps the existing GOAP planner,
+enabling:
+- Persistent beliefs (memory of past events)
+- Long-term desires (personality-driven motivations)
+- Committed intentions (plan persistence)
+
+Main entry points:
+- get_bdi_decision(): Get an AI decision using BDI reasoning
+- reset_bdi_state(): Reset all agent BDI state (on simulation reset)
+"""
+
+from backend.core.bdi.belief import BeliefBase, MemoryEvent
+from backend.core.bdi.desire import Desire, DesireType, DesireManager
+from backend.core.bdi.intention import (
+    Intention,
+    IntentionManager,
+    CommitmentStrategy,
+)
+from backend.core.bdi.bdi_agent import (
+    BDIAgentAI,
+    AIDecision,
+    TradeItem,
+    get_bdi_decision,
+    reset_bdi_state,
+    remove_agent_bdi_state,
+)
+
+__all__ = [
+    # Belief system
+    "BeliefBase",
+    "MemoryEvent",
+    # Desire system
+    "Desire",
+    "DesireType",
+    "DesireManager",
+    # Intention system
+    "Intention",
+    "IntentionManager",
+    "CommitmentStrategy",
+    # BDI Agent
+    "BDIAgentAI",
+    "AIDecision",
+    "TradeItem",
+    # Entry points
+    "get_bdi_decision",
+    "reset_bdi_state",
+    "remove_agent_bdi_state",
+]

backend/core/bdi/bdi_agent.py

@@ -0,0 +1,627 @@
+"""BDI Agent AI that wraps GOAP planning with BDI reasoning.
+
+This module provides the main BDI-based AI decision maker that:
+1. Maintains persistent beliefs about the world
+2. Manages desires based on personality
+3. Commits to intentions (plans) and executes them
+4. Uses GOAP planning to generate action sequences
+
+Performance optimizations:
+- Timeslicing: full BDI cycle only runs periodically
+- Plan persistence: reuses plans across turns
+- Cached belief updates: skips unchanged data
+"""
+
+from dataclasses import dataclass, field
+from typing import Optional, TYPE_CHECKING
+
+from backend.domain.action import ActionType
+from backend.domain.resources import ResourceType
+from backend.domain.personality import get_trade_price_modifier
+
+from backend.core.bdi.belief import BeliefBase
+from backend.core.bdi.desire import DesireManager
+from backend.core.bdi.intention import IntentionManager
+
+from backend.core.goap.planner import GOAPPlanner, ReactivePlanner
+from backend.core.goap.goals import get_all_goals
+from backend.core.goap.actions import get_all_actions
+
+if TYPE_CHECKING:
+    from backend.domain.agent import Agent
+    from backend.core.market import OrderBook
+    from backend.core.goap.goal import Goal
+    from backend.core.goap.action import GOAPAction
+    from backend.core.goap.planner import Plan
+
+
+@dataclass
+class TradeItem:
+    """A single item to buy/sell in a trade."""
+    order_id: str
+    resource_type: ResourceType
+    quantity: int
+    price_per_unit: int
+
+
+@dataclass
+class AIDecision:
+    """A decision made by the AI for an agent."""
+    action: ActionType
+    target_resource: Optional[ResourceType] = None
+    order_id: Optional[str] = None
+    quantity: int = 1
+    price: int = 0
+    reason: str = ""
+    trade_items: list[TradeItem] = field(default_factory=list)
+    adjust_order_id: Optional[str] = None
+    new_price: Optional[int] = None
+
+    # GOAP/BDI-specific fields
+    goal_name: str = ""
+    plan_length: int = 0
+    bdi_info: dict = field(default_factory=dict)
+
+    def to_dict(self) -> dict:
+        return {
+            "action": self.action.value,
+            "target_resource": self.target_resource.value if self.target_resource else None,
+            "order_id": self.order_id,
+            "quantity": self.quantity,
+            "price": self.price,
+            "reason": self.reason,
+            "trade_items": [
+                {
+                    "order_id": t.order_id,
+                    "resource_type": t.resource_type.value,
+                    "quantity": t.quantity,
+                    "price_per_unit": t.price_per_unit,
+                }
+                for t in self.trade_items
+            ],
+            "adjust_order_id": self.adjust_order_id,
+            "new_price": self.new_price,
+            "goal_name": self.goal_name,
+            "plan_length": self.plan_length,
+            "bdi_info": self.bdi_info,
+        }
+
+
+class BDIAgentAI:
+    """BDI-based AI decision maker that wraps GOAP planning.
+    
+    The BDI cycle:
+    1. Update beliefs from sensors (agent state, market)
+    2. Update desires based on beliefs and personality
+    3. Check if current intention should continue
+    4. If needed, generate new plan via GOAP
+    5. Execute next action from intention
+    
+    Performance features:
+    - Timeslicing: full deliberation only every N turns
+    - Plan persistence: reuse plans across turns
+    - Reactive fallback: simple decisions when not deliberating
+    """
+    
+    # Class-level cache for planners (shared across instances)
+    _planner_cache: Optional[GOAPPlanner] = None
+    _reactive_cache: Optional[ReactivePlanner] = None
+    _goals_cache: Optional[list] = None
+    _actions_cache: Optional[list] = None
+    
+    def __init__(
+        self,
+        agent: "Agent",
+        market: "OrderBook",
+        step_in_day: int = 1,
+        day_steps: int = 10,
+        current_turn: int = 0,
+        is_night: bool = False,
+        # Persistent BDI state (passed in for continuity)
+        beliefs: Optional[BeliefBase] = None,
+        desires: Optional[DesireManager] = None,
+        intentions: Optional[IntentionManager] = None,
+    ):
+        self.agent = agent
+        self.market = market
+        self.step_in_day = step_in_day
+        self.day_steps = day_steps
+        self.current_turn = current_turn
+        self.is_night = is_night
+        
+        # Initialize or use existing BDI components
+        self.beliefs = beliefs or BeliefBase()
+        self.desires = desires or DesireManager(agent.personality)
+        self.intentions = intentions or IntentionManager.from_personality(agent.personality)
+        
+        # Update beliefs from current state
+        self.beliefs.update_from_sensors(
+            agent=agent,
+            market=market,
+            step_in_day=step_in_day,
+            day_steps=day_steps,
+            current_turn=current_turn,
+            is_night=is_night,
+        )
+        
+        # Update desires from beliefs
+        self.desires.update_from_beliefs(self.beliefs)
+        
+        # Get cached planners and goals/actions
+        self.planner = self._get_planner()
+        self.reactive_planner = self._get_reactive_planner()
+        self.goals = self._get_goals()
+        self.actions = self._get_actions()
+        
+        # Personality shortcuts
+        self.p = agent.personality
+        self.skills = agent.skills
+    
+    @classmethod
+    def _get_planner(cls) -> GOAPPlanner:
+        """Get cached GOAP planner."""
+        if cls._planner_cache is None:
+            from backend.config import get_config
+            config = get_config()
+            ai_config = config.ai
+            cls._planner_cache = GOAPPlanner(
+                max_iterations=ai_config.goap_max_iterations,
+            )
+        return cls._planner_cache
+    
+    @classmethod
+    def _get_reactive_planner(cls) -> ReactivePlanner:
+        """Get cached reactive planner."""
+        if cls._reactive_cache is None:
+            cls._reactive_cache = ReactivePlanner()
+        return cls._reactive_cache
+    
+    @classmethod
+    def _get_goals(cls) -> list:
+        """Get cached goals list."""
+        if cls._goals_cache is None:
+            cls._goals_cache = get_all_goals()
+        return cls._goals_cache
+    
+    @classmethod
+    def _get_actions(cls) -> list:
+        """Get cached actions list."""
+        if cls._actions_cache is None:
+            cls._actions_cache = get_all_actions()
+        return cls._actions_cache
+    
+    @classmethod
+    def reset_caches(cls) -> None:
+        """Reset all caches (call after config reload)."""
+        cls._planner_cache = None
+        cls._reactive_cache = None
+        cls._goals_cache = None
+        cls._actions_cache = None
+    
+    def should_deliberate(self) -> bool:
+        """Check if this agent should run full BDI deliberation this turn.
+        
+        Timeslicing: not every agent deliberates every turn.
+        Agents are staggered based on their ID hash.
+        """
+        from backend.config import get_config
+        config = get_config()
+        
+        # Get thinking interval from config (default to 1 = every turn)
+        bdi_config = getattr(config, 'bdi', None)
+        thinking_interval = getattr(bdi_config, 'thinking_interval', 1) if bdi_config else 1
+        
+        if thinking_interval <= 1:
+            return True  # Deliberate every turn
+        
+        # Stagger agents across turns
+        agent_hash = hash(self.agent.id) % thinking_interval
+        return (self.current_turn % thinking_interval) == agent_hash
+    
+    def decide(self) -> AIDecision:
+        """Make a decision using BDI reasoning with GOAP planning.
+        
+        Decision flow:
+        1. Night time: mandatory sleep
+        2. Check if should deliberate (timeslicing)
+        3. If deliberating: run full BDI cycle
+        4. If not: continue current intention or reactive fallback
+        """
+        # Night time - mandatory sleep
+        if self.is_night:
+            return AIDecision(
+                action=ActionType.SLEEP,
+                reason="Night time: sleeping",
+                goal_name="Sleep",
+                bdi_info={"mode": "night"},
+            )
+        
+        # Check if we should run full deliberation
+        if self.should_deliberate():
+            return self._deliberate()
+        else:
+            return self._continue_or_react()
+    
+    def _deliberate(self) -> AIDecision:
+        """Run full BDI deliberation cycle."""
+        # Filter goals by desires
+        filtered_goals = self.desires.filter_goals_by_desire(self.goals, self.beliefs)
+        
+        # Check if we should reconsider current intention
+        should_replan = self.intentions.should_reconsider(
+            beliefs=self.beliefs,
+            desire_manager=self.desires,
+            available_goals=filtered_goals,
+        )
+        
+        if not should_replan and self.intentions.has_intention():
+            # Continue with current intention
+            action = self.intentions.get_next_action()
+            if action:
+                return self._convert_to_decision(
+                    goap_action=action,
+                    goal=self.intentions.current_intention.goal,
+                    plan=self.intentions.current_intention.plan,
+                    mode="continue",
+                )
+        
+        # Need to plan for a goal
+        world_state = self.beliefs.to_world_state()
+        
+        plan = self.planner.plan_for_goals(
+            initial_state=world_state,
+            goals=filtered_goals,
+            available_actions=self.actions,
+        )
+        
+        if plan and not plan.is_empty:
+            # Commit to new intention
+            self.intentions.commit_to_plan(
+                goal=plan.goal,
+                plan=plan,
+                current_turn=self.current_turn,
+            )
+            
+            goap_action = plan.first_action
+            return self._convert_to_decision(
+                goap_action=goap_action,
+                goal=plan.goal,
+                plan=plan,
+                mode="new_plan",
+            )
+        
+        # Fallback to reactive planning
+        return self._reactive_fallback()
+    
+    def _continue_or_react(self) -> AIDecision:
+        """Continue current intention or use reactive fallback (no deliberation)."""
+        if self.intentions.has_intention():
+            action = self.intentions.get_next_action()
+            if action:
+                return self._convert_to_decision(
+                    goap_action=action,
+                    goal=self.intentions.current_intention.goal,
+                    plan=self.intentions.current_intention.plan,
+                    mode="timeslice_continue",
+                )
+        
+        # No intention, use reactive fallback
+        return self._reactive_fallback()
+    
+    def _reactive_fallback(self) -> AIDecision:
+        """Use reactive planning when no intention exists."""
+        world_state = self.beliefs.to_world_state()
+        
+        best_action = self.reactive_planner.select_best_action(
+            state=world_state,
+            goals=self.goals,
+            available_actions=self.actions,
+        )
+        
+        if best_action:
+            return self._convert_to_decision(
+                goap_action=best_action,
+                goal=None,
+                plan=None,
+                mode="reactive",
+            )
+        
+        # Ultimate fallback - rest
+        return AIDecision(
+            action=ActionType.REST,
+            reason="No valid action found, resting",
+            bdi_info={"mode": "fallback"},
+        )
+    
+    def _convert_to_decision(
+        self,
+        goap_action: "GOAPAction",
+        goal: Optional["Goal"],
+        plan: Optional["Plan"],
+        mode: str = "deliberate",
+    ) -> AIDecision:
+        """Convert a GOAP action to an AIDecision with proper parameters."""
+        action_type = goap_action.action_type
+        target_resource = goap_action.target_resource
+        
+        # Build reason string
+        if goal:
+            reason = f"{goal.name}: {goap_action.name}"
+        else:
+            reason = f"Reactive: {goap_action.name}"
+        
+        # BDI debug info
+        bdi_info = {
+            "mode": mode,
+            "dominant_desire": self.desires.dominant_desire.value if self.desires.dominant_desire else None,
+            "commitment": self.intentions.commitment_strategy.value,
+            "has_intention": self.intentions.has_intention(),
+        }
+        
+        # Handle different action types
+        if action_type == ActionType.CONSUME:
+            return AIDecision(
+                action=action_type,
+                target_resource=target_resource,
+                reason=reason,
+                goal_name=goal.name if goal else "",
+                plan_length=len(plan.actions) if plan else 0,
+                bdi_info=bdi_info,
+            )
+        
+        elif action_type == ActionType.TRADE:
+            return self._create_trade_decision(goap_action, goal, plan, reason, bdi_info)
+        
+        elif action_type in [ActionType.HUNT, ActionType.GATHER, ActionType.CHOP_WOOD,
+                            ActionType.GET_WATER, ActionType.WEAVE]:
+            return AIDecision(
+                action=action_type,
+                target_resource=target_resource,
+                reason=reason,
+                goal_name=goal.name if goal else "",
+                plan_length=len(plan.actions) if plan else 0,
+                bdi_info=bdi_info,
+            )
+        
+        elif action_type == ActionType.BUILD_FIRE:
+            return AIDecision(
+                action=action_type,
+                target_resource=ResourceType.WOOD,
+                reason=reason,
+                goal_name=goal.name if goal else "",
+                plan_length=len(plan.actions) if plan else 0,
+                bdi_info=bdi_info,
+            )
+        
+        elif action_type in [ActionType.REST, ActionType.SLEEP]:
+            return AIDecision(
+                action=action_type,
+                reason=reason,
+                goal_name=goal.name if goal else "",
+                plan_length=len(plan.actions) if plan else 0,
+                bdi_info=bdi_info,
+            )
+        
+        # Default case
+        return AIDecision(
+            action=action_type,
+            target_resource=target_resource,
+            reason=reason,
+            goal_name=goal.name if goal else "",
+            plan_length=len(plan.actions) if plan else 0,
+            bdi_info=bdi_info,
+        )
+    
+    def _create_trade_decision(
+        self,
+        goap_action: "GOAPAction",
+        goal: Optional["Goal"],
+        plan: Optional["Plan"],
+        reason: str,
+        bdi_info: dict,
+    ) -> AIDecision:
+        """Create a trade decision with actual market parameters."""
+        target_resource = goap_action.target_resource
+        action_name = goap_action.name.lower()
+        
+        if "buy" in action_name:
+            # Find the best order to buy from
+            order = self.market.get_cheapest_order(target_resource)
+            
+            if order and order.seller_id != self.agent.id:
+                # Check trust for this seller
+                trust = self.beliefs.get_trade_trust(order.seller_id)
+                
+                # Skip distrusted sellers if we're picky
+                if trust < -0.5 and self.p.price_sensitivity > 1.2:
+                    # Try next cheapest? For now, fall back to gathering
+                    return self._create_gather_fallback(target_resource, reason, goal, plan, bdi_info)
+                
+                # Calculate quantity to buy
+                can_afford = self.agent.money // max(1, order.price_per_unit)
+                space = self.agent.inventory_space()
+                quantity = min(2, can_afford, space, order.quantity)
+                
+                if quantity > 0:
+                    return AIDecision(
+                        action=ActionType.TRADE,
+                        target_resource=target_resource,
+                        order_id=order.id,
+                        quantity=quantity,
+                        price=order.price_per_unit,
+                        reason=f"{reason} @ {order.price_per_unit}c",
+                        goal_name=goal.name if goal else "",
+                        plan_length=len(plan.actions) if plan else 0,
+                        bdi_info=bdi_info,
+                    )
+            
+            # Can't buy - fallback to gathering
+            return self._create_gather_fallback(target_resource, reason, goal, plan, bdi_info)
+        
+        elif "sell" in action_name:
+            # Create a sell order
+            quantity_available = self.agent.get_resource_count(target_resource)
+            
+            # Calculate minimum to keep
+            min_keep = self._get_min_keep(target_resource)
+            quantity_to_sell = min(3, quantity_available - min_keep)
+            
+            if quantity_to_sell > 0:
+                price = self._calculate_sell_price(target_resource)
+                
+                return AIDecision(
+                    action=ActionType.TRADE,
+                    target_resource=target_resource,
+                    quantity=quantity_to_sell,
+                    price=price,
+                    reason=f"{reason} @ {price}c",
+                    goal_name=goal.name if goal else "",
+                    plan_length=len(plan.actions) if plan else 0,
+                    bdi_info=bdi_info,
+                )
+        
+        # Invalid trade action - rest
+        return AIDecision(
+            action=ActionType.REST,
+            reason="Trade not possible",
+            bdi_info=bdi_info,
+        )
+    
+    def _create_gather_fallback(
+        self,
+        resource_type: ResourceType,
+        reason: str,
+        goal: Optional["Goal"],
+        plan: Optional["Plan"],
+        bdi_info: dict,
+    ) -> AIDecision:
+        """Create a gather action as fallback when buying isn't possible."""
+        action_map = {
+            ResourceType.WATER: ActionType.GET_WATER,
+            ResourceType.BERRIES: ActionType.GATHER,
+            ResourceType.MEAT: ActionType.HUNT,
+            ResourceType.WOOD: ActionType.CHOP_WOOD,
+        }
+        
+        action = action_map.get(resource_type, ActionType.GATHER)
+        
+        return AIDecision(
+            action=action,
+            target_resource=resource_type,
+            reason=f"{reason} (gathering instead)",
+            goal_name=goal.name if goal else "",
+            plan_length=len(plan.actions) if plan else 0,
+            bdi_info=bdi_info,
+        )
+    
+    def _get_min_keep(self, resource_type: ResourceType) -> int:
+        """Get minimum quantity to keep for survival."""
+        # Adjusted by hoarding rate from desires
+        hoarding_mult = 0.5 + self.p.hoarding_rate
+        
+        base_min = {
+            ResourceType.WATER: 2,
+            ResourceType.MEAT: 1,
+            ResourceType.BERRIES: 2,
+            ResourceType.WOOD: 1,
+            ResourceType.HIDE: 0,
+        }
+        
+        return int(base_min.get(resource_type, 1) * hoarding_mult)
+    
+    def _calculate_sell_price(self, resource_type: ResourceType) -> int:
+        """Calculate sell price based on fair value and market conditions."""
+        from backend.core.ai import get_energy_cost
+        from backend.config import get_config
+        
+        config = get_config()
+        economy = getattr(config, 'economy', None)
+        energy_to_money_ratio = getattr(economy, 'energy_to_money_ratio', 150) if economy else 150
+        min_price = getattr(economy, 'min_price', 100) if economy else 100
+        
+        energy_cost = get_energy_cost(resource_type)
+        fair_value = max(min_price, int(energy_cost * energy_to_money_ratio))
+        
+        # Apply trading skill
+        sell_modifier = get_trade_price_modifier(self.skills.trading, is_buying=False)
+        
+        # Get market signal
+        signal = self.market.get_market_signal(resource_type)
+        
+        if signal == "sell":  # Scarcity
+            price = int(fair_value * 1.3 * sell_modifier)
+        elif signal == "hold":
+            price = int(fair_value * sell_modifier)
+        else:  # Surplus
+            cheapest = self.market.get_cheapest_order(resource_type)
+            if cheapest and cheapest.seller_id != self.agent.id:
+                price = max(min_price, cheapest.price_per_unit - 1)
+            else:
+                price = int(fair_value * 0.8 * sell_modifier)
+        
+        return max(min_price, price)
+    
+    def record_action_result(self, success: bool, action_type: str) -> None:
+        """Record the result of an action for learning and intention tracking."""
+        # Update intention
+        self.intentions.advance_intention(success)
+        
+        # Update beliefs/memory
+        if success:
+            self.beliefs.record_successful_action(action_type)
+        else:
+            self.beliefs.record_failed_action(action_type)
+
+
+# Persistent BDI state storage for agents
+_agent_bdi_state: dict[str, tuple[BeliefBase, DesireManager, IntentionManager]] = {}
+
+
+def get_bdi_decision(
+    agent: "Agent",
+    market: "OrderBook",
+    step_in_day: int = 1,
+    day_steps: int = 10,
+    current_turn: int = 0,
+    is_night: bool = False,
+) -> AIDecision:
+    """Get a BDI-based AI decision for an agent.
+    
+    This is the main entry point for the BDI AI system.
+    It maintains persistent BDI state for each agent.
+    """
+    # Get or create persistent BDI state
+    if agent.id not in _agent_bdi_state:
+        beliefs = BeliefBase()
+        desires = DesireManager(agent.personality)
+        intentions = IntentionManager.from_personality(agent.personality)
+        _agent_bdi_state[agent.id] = (beliefs, desires, intentions)
+    else:
+        beliefs, desires, intentions = _agent_bdi_state[agent.id]
+    
+    # Create AI instance with persistent state
+    ai = BDIAgentAI(
+        agent=agent,
+        market=market,
+        step_in_day=step_in_day,
+        day_steps=day_steps,
+        current_turn=current_turn,
+        is_night=is_night,
+        beliefs=beliefs,
+        desires=desires,
+        intentions=intentions,
+    )
+    
+    return ai.decide()
+
+
+def reset_bdi_state() -> None:
+    """Reset all BDI state (call on simulation reset)."""
+    global _agent_bdi_state
+    _agent_bdi_state.clear()
+    BDIAgentAI.reset_caches()
+
+
+def remove_agent_bdi_state(agent_id: str) -> None:
+    """Remove BDI state for a specific agent (call on agent death)."""
+    _agent_bdi_state.pop(agent_id, None)

backend/core/bdi/belief.py

@@ -0,0 +1,401 @@
+"""Belief System for BDI agents.
+
+The BeliefBase maintains persistent state about the world from the agent's
+perspective, including:
+- Current sensory data (vitals, inventory, market)
+- Memory of past events (failed trades, good hunting spots)
+- Dirty flags for efficient updates
+
+This replaces the transient WorldState creation with a persistent belief system.
+"""
+
+from dataclasses import dataclass, field
+from typing import TYPE_CHECKING, Optional
+from collections import deque
+
+if TYPE_CHECKING:
+    from backend.domain.agent import Agent
+    from backend.core.market import OrderBook
+
+
+@dataclass
+class MemoryEvent:
+    """A remembered event that may influence future decisions."""
+    event_type: str  # "trade_failed", "hunt_failed", "good_deal", etc.
+    turn: int
+    data: dict = field(default_factory=dict)
+    relevance: float = 1.0  # Decays over time
+
+
+@dataclass
+class BeliefBase:
+    """Persistent belief system for a BDI agent.
+    
+    Maintains both current perceptions and memories of past events.
+    Uses dirty flags to avoid unnecessary recomputation.
+    """
+    
+    # Current perception state (cached WorldState fields)
+    thirst_pct: float = 1.0
+    hunger_pct: float = 1.0
+    heat_pct: float = 1.0
+    energy_pct: float = 1.0
+    
+    # Resource counts
+    water_count: int = 0
+    food_count: int = 0
+    meat_count: int = 0
+    berries_count: int = 0
+    wood_count: int = 0
+    hide_count: int = 0
+    
+    # Inventory state
+    has_clothes: bool = False
+    inventory_space: int = 0
+    inventory_full: bool = False
+    
+    # Economic state
+    money: int = 0
+    is_wealthy: bool = False
+    
+    # Market beliefs (what we believe about market availability)
+    can_buy_water: bool = False
+    can_buy_food: bool = False
+    can_buy_meat: bool = False
+    can_buy_berries: bool = False
+    can_buy_wood: bool = False
+    water_market_price: int = 0
+    food_market_price: int = 0
+    wood_market_price: int = 0
+    
+    # Time beliefs
+    is_night: bool = False
+    is_evening: bool = False
+    step_in_day: int = 0
+    day_steps: int = 10
+    current_turn: int = 0
+    
+    # Personality (cached from agent, rarely changes)
+    wealth_desire: float = 0.5
+    hoarding_rate: float = 0.5
+    risk_tolerance: float = 0.5
+    market_affinity: float = 0.5
+    is_trader: bool = False
+    gather_preference: float = 1.0
+    hunt_preference: float = 1.0
+    trade_preference: float = 1.0
+    
+    # Skills
+    hunting_skill: float = 1.0
+    gathering_skill: float = 1.0
+    trading_skill: float = 1.0
+    
+    # Thresholds (from config)
+    critical_threshold: float = 0.25
+    low_threshold: float = 0.45
+    
+    # Calculated urgencies
+    thirst_urgency: float = 0.0
+    hunger_urgency: float = 0.0
+    heat_urgency: float = 0.0
+    energy_urgency: float = 0.0
+    
+    # === BDI Extensions ===
+    
+    # Memory system - stores past events that influence decisions
+    memories: deque = field(default_factory=lambda: deque(maxlen=50))
+    
+    # Track failed actions for this resource type (turn -> count)
+    failed_hunts: int = 0
+    failed_gathers: int = 0
+    failed_trades: int = 0
+    
+    # Track successful trades with agents (agent_id -> positive count)
+    trusted_traders: dict = field(default_factory=dict)
+    # Track failed trades with agents (agent_id -> negative count)
+    distrusted_traders: dict = field(default_factory=dict)
+    
+    # Dirty flags for optimization
+    _vitals_dirty: bool = True
+    _inventory_dirty: bool = True
+    _market_dirty: bool = True
+    _last_update_turn: int = -1
+    
+    def update_from_sensors(
+        self,
+        agent: "Agent",
+        market: "OrderBook",
+        step_in_day: int = 1,
+        day_steps: int = 10,
+        current_turn: int = 0,
+        is_night: bool = False,
+    ) -> None:
+        """Update beliefs from current agent and market state.
+        
+        Uses dirty flags to skip unchanged data.
+        """
+        from backend.domain.resources import ResourceType
+        from backend.config import get_config
+        
+        self.current_turn = current_turn
+        self.step_in_day = step_in_day
+        self.day_steps = day_steps
+        self.is_night = is_night
+        self.is_evening = step_in_day >= day_steps - 2
+        
+        # Always update vitals (they change every turn)
+        stats = agent.stats
+        self.thirst_pct = stats.thirst / stats.MAX_THIRST
+        self.hunger_pct = stats.hunger / stats.MAX_HUNGER
+        self.heat_pct = stats.heat / stats.MAX_HEAT
+        self.energy_pct = stats.energy / stats.MAX_ENERGY
+        
+        # Update inventory
+        self.water_count = agent.get_resource_count(ResourceType.WATER)
+        self.meat_count = agent.get_resource_count(ResourceType.MEAT)
+        self.berries_count = agent.get_resource_count(ResourceType.BERRIES)
+        self.wood_count = agent.get_resource_count(ResourceType.WOOD)
+        self.hide_count = agent.get_resource_count(ResourceType.HIDE)
+        self.food_count = self.meat_count + self.berries_count
+        
+        self.has_clothes = agent.has_clothes()
+        self.inventory_space = agent.inventory_space()
+        self.inventory_full = agent.inventory_full()
+        self.money = agent.money
+        
+        # Update personality (cached, rarely changes)
+        self.wealth_desire = agent.personality.wealth_desire
+        self.hoarding_rate = agent.personality.hoarding_rate
+        self.risk_tolerance = agent.personality.risk_tolerance
+        self.market_affinity = agent.personality.market_affinity
+        self.gather_preference = agent.personality.gather_preference
+        self.hunt_preference = agent.personality.hunt_preference
+        self.trade_preference = agent.personality.trade_preference
+        
+        # Skills
+        self.hunting_skill = agent.skills.hunting
+        self.gathering_skill = agent.skills.gathering
+        self.trading_skill = agent.skills.trading
+        
+        # Market availability
+        def get_market_info(resource_type: ResourceType) -> tuple[bool, int]:
+            order = market.get_cheapest_order(resource_type)
+            if order and order.seller_id != agent.id and agent.money >= order.price_per_unit:
+                return True, order.price_per_unit
+            return False, 0
+        
+        self.can_buy_water, self.water_market_price = get_market_info(ResourceType.WATER)
+        self.can_buy_meat, meat_price = get_market_info(ResourceType.MEAT)
+        self.can_buy_berries, berries_price = get_market_info(ResourceType.BERRIES)
+        self.can_buy_wood, self.wood_market_price = get_market_info(ResourceType.WOOD)
+        
+        self.can_buy_food = self.can_buy_meat or self.can_buy_berries
+        food_price = min(
+            meat_price if self.can_buy_meat else float('inf'),
+            berries_price if self.can_buy_berries else float('inf')
+        )
+        self.food_market_price = int(food_price) if food_price != float('inf') else 0
+        
+        # Wealth calculation
+        config = get_config()
+        economy_config = getattr(config, 'economy', None)
+        min_wealth_target = getattr(economy_config, 'min_wealth_target', 5000) if economy_config else 5000
+        wealth_target = int(min_wealth_target * (0.5 + self.wealth_desire))
+        self.is_wealthy = self.money >= wealth_target
+        
+        # Trader check
+        self.is_trader = self.trade_preference > 1.3 and self.market_affinity > 0.5
+        
+        # Config thresholds
+        agent_config = config.agent_stats
+        self.critical_threshold = agent_config.critical_threshold
+        self.low_threshold = 0.45
+        
+        # Calculate urgencies
+        self._calculate_urgencies()
+        
+        # Decay old memories
+        self._decay_memories()
+        
+        self._last_update_turn = current_turn
+    
+    def _calculate_urgencies(self) -> None:
+        """Calculate urgency values for each vital stat."""
+        def calc_urgency(pct: float, critical: float, low: float) -> float:
+            if pct >= low:
+                return 0.0
+            elif pct >= critical:
+                return 1.0 - (pct - critical) / (low - critical)
+            else:
+                return 1.0 + (critical - pct) / critical * 2.0
+        
+        self.thirst_urgency = calc_urgency(self.thirst_pct, self.critical_threshold, self.low_threshold)
+        self.hunger_urgency = calc_urgency(self.hunger_pct, self.critical_threshold, self.low_threshold)
+        self.heat_urgency = calc_urgency(self.heat_pct, self.critical_threshold, self.low_threshold)
+        
+        if self.energy_pct < 0.25:
+            self.energy_urgency = 2.0
+        elif self.energy_pct < 0.40:
+            self.energy_urgency = 1.0
+        else:
+            self.energy_urgency = 0.0
+    
+    def _decay_memories(self) -> None:
+        """Decay memory relevance over time."""
+        for memory in self.memories:
+            age = self.current_turn - memory.turn
+            # Memories decay exponentially with age
+            memory.relevance = max(0.1, 1.0 / (1.0 + age * 0.1))
+    
+    def add_memory(self, event_type: str, data: dict = None) -> None:
+        """Add a new memory event."""
+        self.memories.append(MemoryEvent(
+            event_type=event_type,
+            turn=self.current_turn,
+            data=data or {},
+            relevance=1.0,
+        ))
+    
+    def record_failed_action(self, action_type: str) -> None:
+        """Record a failed action for learning."""
+        if action_type == "hunt":
+            self.failed_hunts += 1
+            self.add_memory("hunt_failed")
+        elif action_type == "gather":
+            self.failed_gathers += 1
+            self.add_memory("gather_failed")
+        elif action_type == "trade":
+            self.failed_trades += 1
+            self.add_memory("trade_failed")
+    
+    def record_successful_action(self, action_type: str) -> None:
+        """Record a successful action, reducing failure counts."""
+        if action_type == "hunt":
+            self.failed_hunts = max(0, self.failed_hunts - 1)
+        elif action_type == "gather":
+            self.failed_gathers = max(0, self.failed_gathers - 1)
+        elif action_type == "trade":
+            self.failed_trades = max(0, self.failed_trades - 1)
+    
+    def record_trade_partner(self, partner_id: str, success: bool) -> None:
+        """Track trade relationship with another agent."""
+        if success:
+            self.trusted_traders[partner_id] = self.trusted_traders.get(partner_id, 0) + 1
+            # Reduce distrust if they've been bad before
+            if partner_id in self.distrusted_traders:
+                self.distrusted_traders[partner_id] = max(0, self.distrusted_traders[partner_id] - 1)
+        else:
+            self.distrusted_traders[partner_id] = self.distrusted_traders.get(partner_id, 0) + 1
+    
+    def get_trade_trust(self, partner_id: str) -> float:
+        """Get trust level for a trade partner (-1 to 1)."""
+        trust = self.trusted_traders.get(partner_id, 0)
+        distrust = self.distrusted_traders.get(partner_id, 0)
+        total = trust + distrust
+        if total == 0:
+            return 0.0  # Unknown partner
+        return (trust - distrust) / total
+    
+    def has_critical_need(self) -> bool:
+        """Check if any vital stat is critical (requires immediate attention)."""
+        return (
+            self.thirst_urgency >= 2.0 or
+            self.hunger_urgency >= 2.0 or
+            self.heat_urgency >= 2.0 or
+            self.energy_urgency >= 2.0
+        )
+    
+    def get_most_urgent_need(self) -> Optional[str]:
+        """Get the most urgent vital need, if any."""
+        urgencies = {
+            "thirst": self.thirst_urgency,
+            "hunger": self.hunger_urgency,
+            "heat": self.heat_urgency,
+            "energy": self.energy_urgency,
+        }
+        max_urgency = max(urgencies.values())
+        if max_urgency < 0.5:
+            return None
+        return max(urgencies, key=urgencies.get)
+    
+    def to_world_state(self):
+        """Convert beliefs to a WorldState for GOAP planner compatibility."""
+        from backend.core.goap.world_state import WorldState
+
+        return WorldState(
+            thirst_pct=self.thirst_pct,
+            hunger_pct=self.hunger_pct,
+            heat_pct=self.heat_pct,
+            energy_pct=self.energy_pct,
+            water_count=self.water_count,
+            food_count=self.food_count,
+            meat_count=self.meat_count,
+            berries_count=self.berries_count,
+            wood_count=self.wood_count,
+            hide_count=self.hide_count,
+            has_clothes=self.has_clothes,
+            inventory_space=self.inventory_space,
+            inventory_full=self.inventory_full,
+            money=self.money,
+            is_wealthy=self.is_wealthy,
+            can_buy_water=self.can_buy_water,
+            can_buy_food=self.can_buy_food,
+            can_buy_meat=self.can_buy_meat,
+            can_buy_berries=self.can_buy_berries,
+            can_buy_wood=self.can_buy_wood,
+            water_market_price=self.water_market_price,
+            food_market_price=self.food_market_price,
+            wood_market_price=self.wood_market_price,
+            is_night=self.is_night,
+            is_evening=self.is_evening,
+            step_in_day=self.step_in_day,
+            day_steps=self.day_steps,
+            wealth_desire=self.wealth_desire,
+            hoarding_rate=self.hoarding_rate,
+            risk_tolerance=self.risk_tolerance,
+            market_affinity=self.market_affinity,
+            is_trader=self.is_trader,
+            gather_preference=self.gather_preference,
+            hunt_preference=self.hunt_preference,
+            trade_preference=self.trade_preference,
+            hunting_skill=self.hunting_skill,
+            gathering_skill=self.gathering_skill,
+            trading_skill=self.trading_skill,
+            critical_threshold=self.critical_threshold,
+            low_threshold=self.low_threshold,
+        )
+    
+    def to_dict(self) -> dict:
+        """Convert to dictionary for debugging/logging."""
+        return {
+            "vitals": {
+                "thirst": round(self.thirst_pct, 2),
+                "hunger": round(self.hunger_pct, 2),
+                "heat": round(self.heat_pct, 2),
+                "energy": round(self.energy_pct, 2),
+            },
+            "urgencies": {
+                "thirst": round(self.thirst_urgency, 2),
+                "hunger": round(self.hunger_urgency, 2),
+                "heat": round(self.heat_urgency, 2),
+                "energy": round(self.energy_urgency, 2),
+            },
+            "inventory": {
+                "water": self.water_count,
+                "meat": self.meat_count,
+                "berries": self.berries_count,
+                "wood": self.wood_count,
+                "hide": self.hide_count,
+                "space": self.inventory_space,
+            },
+            "economy": {
+                "money": self.money,
+                "is_wealthy": self.is_wealthy,
+            },
+            "memory": {
+                "failed_hunts": self.failed_hunts,
+                "failed_gathers": self.failed_gathers,
+                "failed_trades": self.failed_trades,
+                "memory_count": len(self.memories),
+            },
+        }

backend/core/bdi/desire.py

@@ -0,0 +1,347 @@
+"""Desire System for BDI agents.
+
+Desires represent high-level, long-term motivations that persist across turns.
+Unlike GOAP goals (which are immediate targets), desires are personality-driven
+and influence which goals get activated.
+
+Key concepts:
+- Desires are weighted by personality traits
+- Desires can be satisfied temporarily but return
+- Desires influence goal selection, not direct actions
+"""
+
+from dataclasses import dataclass
+from enum import Enum
+from typing import TYPE_CHECKING, Optional
+
+if TYPE_CHECKING:
+    from backend.core.bdi.belief import BeliefBase
+    from backend.domain.personality import PersonalityTraits
+    from backend.core.goap.goal import Goal
+
+
+class DesireType(Enum):
+    """Types of high-level desires."""
+    SURVIVAL = "survival"           # Stay alive (thirst, hunger, heat, energy)
+    ACCUMULATE_WEALTH = "wealth"    # Build up money reserves
+    STOCK_RESOURCES = "stock"       # Hoard resources for security
+    MASTER_PROFESSION = "mastery"   # Improve skills in preferred activity
+    SOCIAL_STANDING = "social"      # Gain reputation through trade
+    COMFORT = "comfort"             # Maintain clothes, warmth, rest
+
+
+@dataclass
+class Desire:
+    """A high-level motivation that influences goal selection.
+    
+    Desires have:
+    - A base intensity (how strongly the agent wants this)
+    - A satisfaction level (0-1, how fulfilled is this desire currently)
+    - A personality weight (how much this agent's personality cares)
+    """
+    desire_type: DesireType
+    base_intensity: float = 1.0      # Base importance (0-2)
+    satisfaction: float = 0.5        # Current fulfillment (0-1)
+    personality_weight: float = 1.0  # Multiplier from personality
+    
+    # Persistence tracking
+    turns_pursued: int = 0           # How long we've been chasing this
+    turns_since_progress: int = 0    # Turns since we made progress
+    
+    @property
+    def effective_intensity(self) -> float:
+        """Calculate current desire intensity considering satisfaction."""
+        # Desire is stronger when unsatisfied
+        unsatisfied = 1.0 - self.satisfaction
+        # Apply personality weight
+        intensity = self.base_intensity * self.personality_weight * (0.5 + unsatisfied)
+        # Reduce intensity if pursued too long without progress (boredom)
+        if self.turns_since_progress > 10:
+            boredom_factor = max(0.3, 1.0 - (self.turns_since_progress - 10) * 0.05)
+            intensity *= boredom_factor
+        return intensity
+    
+    def update_satisfaction(self, new_satisfaction: float) -> None:
+        """Update satisfaction level and track progress."""
+        if new_satisfaction > self.satisfaction:
+            self.turns_since_progress = 0  # Made progress!
+        else:
+            self.turns_since_progress += 1
+        self.satisfaction = max(0.0, min(1.0, new_satisfaction))
+        self.turns_pursued += 1
+    
+    def reset_pursuit(self) -> None:
+        """Reset pursuit tracking (when switching to different desire)."""
+        self.turns_pursued = 0
+        self.turns_since_progress = 0
+
+
+class DesireManager:
+    """Manages an agent's desires and converts them to active goals.
+    
+    The DesireManager:
+    1. Maintains a set of desires weighted by personality
+    2. Updates desire satisfaction based on beliefs
+    3. Selects which desires should drive current goals
+    4. Filters/prioritizes GOAP goals based on active desires
+    """
+    
+    def __init__(self, personality: "PersonalityTraits"):
+        """Initialize desires based on personality."""
+        self.desires: dict[DesireType, Desire] = {}
+        self._init_desires(personality)
+        
+        # Track which desire is currently dominant
+        self.dominant_desire: Optional[DesireType] = None
+        self.stubbornness: float = 0.5  # How reluctant to switch desires
+    
+    def _init_desires(self, personality: "PersonalityTraits") -> None:
+        """Initialize desires with personality-based weights."""
+        # Survival - everyone has this, but risk-averse agents weight it higher
+        self.desires[DesireType.SURVIVAL] = Desire(
+            desire_type=DesireType.SURVIVAL,
+            base_intensity=2.0,  # Always high base
+            personality_weight=1.5 - personality.risk_tolerance * 0.5,
+        )
+        
+        # Accumulate Wealth - driven by wealth_desire
+        self.desires[DesireType.ACCUMULATE_WEALTH] = Desire(
+            desire_type=DesireType.ACCUMULATE_WEALTH,
+            base_intensity=1.0,
+            personality_weight=0.5 + personality.wealth_desire,
+        )
+        
+        # Stock Resources - driven by hoarding_rate
+        self.desires[DesireType.STOCK_RESOURCES] = Desire(
+            desire_type=DesireType.STOCK_RESOURCES,
+            base_intensity=0.8,
+            personality_weight=0.5 + personality.hoarding_rate,
+        )
+        
+        # Master Profession - driven by strongest preference
+        max_pref = max(
+            personality.hunt_preference,
+            personality.gather_preference,
+            personality.trade_preference,
+        )
+        self.desires[DesireType.MASTER_PROFESSION] = Desire(
+            desire_type=DesireType.MASTER_PROFESSION,
+            base_intensity=0.6,
+            personality_weight=max_pref * 0.5,
+        )
+        
+        # Social Standing - driven by market_affinity and trade_preference
+        self.desires[DesireType.SOCIAL_STANDING] = Desire(
+            desire_type=DesireType.SOCIAL_STANDING,
+            base_intensity=0.5,
+            personality_weight=personality.market_affinity * personality.trade_preference * 0.5,
+        )
+        
+        # Comfort - moderate for everyone
+        self.desires[DesireType.COMFORT] = Desire(
+            desire_type=DesireType.COMFORT,
+            base_intensity=0.7,
+            personality_weight=1.0 - personality.risk_tolerance * 0.3,
+        )
+        
+        # Calculate stubbornness from personality
+        # High hoarding + low risk tolerance = stubborn
+        self.stubbornness = (personality.hoarding_rate + (1.0 - personality.risk_tolerance)) / 2
+    
+    def update_from_beliefs(self, beliefs: "BeliefBase") -> None:
+        """Update desire satisfaction based on current beliefs."""
+        # Survival satisfaction based on vitals
+        min_vital = min(
+            beliefs.thirst_pct,
+            beliefs.hunger_pct,
+            beliefs.heat_pct,
+            beliefs.energy_pct,
+        )
+        self.desires[DesireType.SURVIVAL].update_satisfaction(min_vital)
+        
+        # Wealth satisfaction based on money relative to wealthy threshold
+        if beliefs.is_wealthy:
+            self.desires[DesireType.ACCUMULATE_WEALTH].update_satisfaction(0.8)
+        else:
+            # Scale satisfaction by money (rough approximation)
+            wealth_sat = min(1.0, beliefs.money / 10000)
+            self.desires[DesireType.ACCUMULATE_WEALTH].update_satisfaction(wealth_sat)
+        
+        # Stock satisfaction based on inventory
+        total_resources = (
+            beliefs.water_count +
+            beliefs.food_count +
+            beliefs.wood_count
+        )
+        stock_sat = min(1.0, total_resources / 15)  # 15 resources = satisfied
+        self.desires[DesireType.STOCK_RESOURCES].update_satisfaction(stock_sat)
+        
+        # Mastery satisfaction based on skill levels
+        max_skill = max(
+            beliefs.hunting_skill,
+            beliefs.gathering_skill,
+            beliefs.trading_skill,
+        )
+        mastery_sat = (max_skill - 1.0) / 1.0  # 0 at skill 1.0, 1 at skill 2.0
+        self.desires[DesireType.MASTER_PROFESSION].update_satisfaction(max(0, mastery_sat))
+        
+        # Social satisfaction - harder to measure, use trade success
+        trade_success = max(0, 5 - beliefs.failed_trades) / 5
+        self.desires[DesireType.SOCIAL_STANDING].update_satisfaction(trade_success)
+        
+        # Comfort satisfaction
+        comfort_factors = [
+            beliefs.energy_pct,
+            beliefs.heat_pct,
+            1.0 if beliefs.has_clothes else 0.5,
+        ]
+        comfort_sat = sum(comfort_factors) / len(comfort_factors)
+        self.desires[DesireType.COMFORT].update_satisfaction(comfort_sat)
+    
+    def get_active_desires(self) -> list[Desire]:
+        """Get desires sorted by effective intensity (highest first)."""
+        return sorted(
+            self.desires.values(),
+            key=lambda d: d.effective_intensity,
+            reverse=True,
+        )
+    
+    def should_switch_desire(self, new_dominant: DesireType) -> bool:
+        """Determine if we should switch dominant desire.
+        
+        Stubborn agents require a larger intensity difference to switch.
+        """
+        if self.dominant_desire is None:
+            return True
+        
+        if new_dominant == self.dominant_desire:
+            return False
+        
+        current = self.desires[self.dominant_desire]
+        new = self.desires[new_dominant]
+        
+        # Calculate required intensity difference based on stubbornness
+        # Stubbornness 0.5 = need 20% more intensity to switch
+        # Stubbornness 1.0 = need 50% more intensity to switch
+        required_difference = 1.0 + self.stubbornness * 0.5
+        
+        return new.effective_intensity > current.effective_intensity * required_difference
+    
+    def update_dominant_desire(self) -> DesireType:
+        """Update and return the currently dominant desire."""
+        active = self.get_active_desires()
+        if not active:
+            return DesireType.SURVIVAL
+        
+        top_desire = active[0].desire_type
+        
+        if self.should_switch_desire(top_desire):
+            # Reset pursuit on old desire if switching
+            if self.dominant_desire and self.dominant_desire != top_desire:
+                self.desires[self.dominant_desire].reset_pursuit()
+            self.dominant_desire = top_desire
+        
+        return self.dominant_desire
+    
+    def filter_goals_by_desire(
+        self,
+        goals: list["Goal"],
+        beliefs: "BeliefBase",
+    ) -> list["Goal"]:
+        """Filter and re-prioritize goals based on active desires.
+        
+        This is where desires influence GOAP goal selection:
+        - Goals aligned with dominant desire get priority boost
+        - Goals conflicting with desires get reduced priority
+        """
+        # Update dominant desire
+        dominant = self.update_dominant_desire()
+        
+        # Get world state for goal priority calculation
+        world_state = beliefs.to_world_state()
+        
+        # Calculate goal priorities with desire modifiers
+        goal_priorities: list[tuple["Goal", float]] = []
+        
+        for goal in goals:
+            base_priority = goal.get_priority(world_state)
+            
+            # Apply desire-based modifiers
+            modifier = self._get_goal_desire_modifier(goal.name, dominant, beliefs)
+            adjusted_priority = base_priority * modifier
+            
+            goal_priorities.append((goal, adjusted_priority))
+        
+        # Sort by adjusted priority
+        goal_priorities.sort(key=lambda x: x[1], reverse=True)
+        
+        # Return goals (the priorities will be recalculated by planner,
+        # but we've filtered/ordered them by our preferences)
+        return [g for g, _ in goal_priorities]
+    
+    def _get_goal_desire_modifier(
+        self,
+        goal_name: str,
+        dominant_desire: DesireType,
+        beliefs: "BeliefBase",
+    ) -> float:
+        """Get priority modifier for a goal based on dominant desire."""
+        goal_lower = goal_name.lower()
+        
+        # Map goals to desires they serve
+        goal_desire_map = {
+            # Survival goals
+            "satisfy thirst": DesireType.SURVIVAL,
+            "satisfy hunger": DesireType.SURVIVAL,
+            "maintain heat": DesireType.SURVIVAL,
+            "restore energy": DesireType.SURVIVAL,
+            # Wealth goals
+            "build wealth": DesireType.ACCUMULATE_WEALTH,
+            "sell excess": DesireType.ACCUMULATE_WEALTH,
+            "find deals": DesireType.ACCUMULATE_WEALTH,
+            "trader arbitrage": DesireType.ACCUMULATE_WEALTH,
+            # Stock goals
+            "stock water": DesireType.STOCK_RESOURCES,
+            "stock food": DesireType.STOCK_RESOURCES,
+            "stock wood": DesireType.STOCK_RESOURCES,
+            # Comfort goals
+            "get clothes": DesireType.COMFORT,
+            "sleep": DesireType.COMFORT,
+        }
+        
+        # Find which desire this goal serves
+        goal_desire = None
+        for gn, desire in goal_desire_map.items():
+            if gn in goal_lower:
+                goal_desire = desire
+                break
+        
+        if goal_desire is None:
+            return 1.0  # No modifier for unknown goals
+        
+        # Boost if aligned with dominant desire
+        if goal_desire == dominant_desire:
+            return 1.3 + self.desires[dominant_desire].effective_intensity * 0.2
+        
+        # Survival always gets a baseline boost if critical
+        if goal_desire == DesireType.SURVIVAL and beliefs.has_critical_need():
+            return 2.0  # Critical needs override other desires
+        
+        # Slight reduction for non-aligned goals
+        return 0.9
+    
+    def to_dict(self) -> dict:
+        """Convert to dictionary for debugging/logging."""
+        return {
+            "dominant_desire": self.dominant_desire.value if self.dominant_desire else None,
+            "stubbornness": round(self.stubbornness, 2),
+            "desires": {
+                d.desire_type.value: {
+                    "intensity": round(d.effective_intensity, 2),
+                    "satisfaction": round(d.satisfaction, 2),
+                    "personality_weight": round(d.personality_weight, 2),
+                    "turns_pursued": d.turns_pursued,
+                }
+                for d in self.desires.values()
+            },
+        }

backend/core/bdi/intention.py

@@ -0,0 +1,289 @@
+"""Intention System for BDI agents.
+
+Intentions represent committed plans that the agent is executing.
+Unlike desires (motivations) or goals (targets), intentions are
+concrete action sequences the agent has decided to pursue.
+
+Key concepts:
+- Intention persistence: agents stick to plans unless interrupted
+- Commitment strategies: different levels of plan commitment
+- Plan monitoring: detecting when a plan becomes invalid
+"""
+
+from dataclasses import dataclass
+from enum import Enum
+from typing import TYPE_CHECKING, Optional
+
+if TYPE_CHECKING:
+    from backend.core.bdi.belief import BeliefBase
+    from backend.core.bdi.desire import DesireManager
+    from backend.core.goap.goal import Goal
+    from backend.core.goap.action import GOAPAction
+    from backend.core.goap.planner import Plan
+
+
+class CommitmentStrategy(Enum):
+    """How strongly an agent commits to their current intention."""
+    REACTIVE = "reactive"       # Replan every turn (no commitment)
+    CAUTIOUS = "cautious"       # Replan if priorities shift significantly
+    DETERMINED = "determined"   # Stick to plan unless it becomes impossible
+    STUBBORN = "stubborn"       # Only abandon for critical interrupts
+
+
+@dataclass
+class Intention:
+    """A committed plan that the agent is executing.
+    
+    Tracks the goal, plan, and execution progress.
+    """
+    goal: "Goal"                           # The goal we're pursuing
+    plan: "Plan"                           # The plan to achieve it
+    start_turn: int                        # When we committed
+    actions_completed: int = 0             # How many actions done
+    last_action_success: bool = True       # Did last action succeed?
+    consecutive_failures: int = 0          # Failures in a row
+    
+    @property
+    def current_action(self) -> Optional["GOAPAction"]:
+        """Get the next action to execute."""
+        if self.plan.is_empty:
+            return None
+        remaining = self.plan.actions[self.actions_completed:]
+        return remaining[0] if remaining else None
+    
+    @property
+    def is_complete(self) -> bool:
+        """Check if the plan has been fully executed."""
+        return self.actions_completed >= len(self.plan.actions)
+    
+    @property
+    def remaining_actions(self) -> int:
+        """Number of actions left in the plan."""
+        return max(0, len(self.plan.actions) - self.actions_completed)
+    
+    def advance(self, success: bool) -> None:
+        """Mark current action as executed and advance."""
+        self.last_action_success = success
+        if success:
+            self.actions_completed += 1
+            self.consecutive_failures = 0
+        else:
+            self.consecutive_failures += 1
+
+
+class IntentionManager:
+    """Manages the agent's current intention and commitment.
+    
+    Responsibilities:
+    - Maintain the current intention (goal + plan)
+    - Decide when to continue vs. replan
+    - Handle plan failure and recovery
+    """
+    
+    def __init__(self, commitment_strategy: CommitmentStrategy = CommitmentStrategy.CAUTIOUS):
+        self.current_intention: Optional[Intention] = None
+        self.commitment_strategy = commitment_strategy
+        
+        # Tracking
+        self.intentions_completed: int = 0
+        self.intentions_abandoned: int = 0
+        self.total_actions_executed: int = 0
+        
+        # Thresholds for reconsideration
+        self.max_consecutive_failures: int = 2
+        self.priority_switch_threshold: float = 1.5  # New goal must be 1.5x priority
+    
+    @classmethod
+    def from_personality(cls, personality) -> "IntentionManager":
+        """Create an IntentionManager with commitment strategy based on personality."""
+        # Derive commitment from personality traits
+        # High hoarding + low risk tolerance = stubborn
+        commitment_score = (
+            personality.hoarding_rate * 0.4 +
+            (1.0 - personality.risk_tolerance) * 0.4 +
+            (1.0 - personality.market_affinity) * 0.2
+        )
+        
+        if commitment_score > 0.7:
+            strategy = CommitmentStrategy.STUBBORN
+        elif commitment_score > 0.5:
+            strategy = CommitmentStrategy.DETERMINED
+        elif commitment_score > 0.3:
+            strategy = CommitmentStrategy.CAUTIOUS
+        else:
+            strategy = CommitmentStrategy.REACTIVE
+        
+        return cls(commitment_strategy=strategy)
+    
+    def has_intention(self) -> bool:
+        """Check if we have an active intention."""
+        return (
+            self.current_intention is not None and
+            not self.current_intention.is_complete
+        )
+    
+    def get_next_action(self) -> Optional["GOAPAction"]:
+        """Get the next action from current intention."""
+        if not self.has_intention():
+            return None
+        return self.current_intention.current_action
+    
+    def should_reconsider(
+        self,
+        beliefs: "BeliefBase",
+        desire_manager: "DesireManager",
+        available_goals: list["Goal"],
+    ) -> bool:
+        """Determine if we should reconsider our current intention.
+        
+        This implements the commitment strategy logic.
+        """
+        # No intention = definitely need to plan
+        if not self.has_intention():
+            return True
+        
+        intention = self.current_intention
+        
+        # Check for critical interrupts (always reconsider)
+        if beliefs.has_critical_need():
+            # But only if current intention isn't already addressing it
+            urgent_need = beliefs.get_most_urgent_need()
+            if not self._intention_addresses_need(intention, urgent_need):
+                return True
+        
+        # Check for too many failures
+        if intention.consecutive_failures >= self.max_consecutive_failures:
+            return True
+        
+        # Strategy-specific logic
+        if self.commitment_strategy == CommitmentStrategy.REACTIVE:
+            return True  # Always replan
+        
+        elif self.commitment_strategy == CommitmentStrategy.CAUTIOUS:
+            # Reconsider if a significantly better goal is available
+            return self._better_goal_available(
+                beliefs, desire_manager, available_goals,
+                threshold=self.priority_switch_threshold
+            )
+        
+        elif self.commitment_strategy == CommitmentStrategy.DETERMINED:
+            # Only reconsider for much better goals or if plan is failing
+            return (
+                intention.consecutive_failures > 0 or
+                self._better_goal_available(
+                    beliefs, desire_manager, available_goals,
+                    threshold=2.0  # Need 2x priority to switch
+                )
+            )
+        
+        elif self.commitment_strategy == CommitmentStrategy.STUBBORN:
+            # Only abandon for critical needs or impossible plans
+            return (
+                beliefs.has_critical_need() or
+                intention.consecutive_failures >= self.max_consecutive_failures
+            )
+        
+        return False
+    
+    def _intention_addresses_need(self, intention: Intention, need: str) -> bool:
+        """Check if current intention addresses a vital need."""
+        if not intention or not need:
+            return False
+        
+        goal_name = intention.goal.name.lower()
+        need_map = {
+            "thirst": "thirst",
+            "hunger": "hunger",
+            "heat": "heat",
+            "energy": "energy",
+        }
+        
+        return need_map.get(need, "") in goal_name
+    
+    def _better_goal_available(
+        self,
+        beliefs: "BeliefBase",
+        desire_manager: "DesireManager",
+        available_goals: list["Goal"],
+        threshold: float,
+    ) -> bool:
+        """Check if there's a significantly better goal available."""
+        if not self.has_intention():
+            return True
+        
+        current_goal = self.current_intention.goal
+        world_state = beliefs.to_world_state()
+        current_priority = current_goal.get_priority(world_state)
+        
+        # Get desire-filtered goals
+        filtered_goals = desire_manager.filter_goals_by_desire(available_goals, beliefs)
+        
+        for goal in filtered_goals:
+            if goal.name == current_goal.name:
+                continue
+            
+            goal_priority = goal.get_priority(world_state)
+            if goal_priority > current_priority * threshold:
+                return True
+        
+        return False
+    
+    def commit_to_plan(self, goal: "Goal", plan: "Plan", current_turn: int) -> None:
+        """Commit to a new intention."""
+        # Track abandoned intention
+        if self.has_intention():
+            self.intentions_abandoned += 1
+        
+        self.current_intention = Intention(
+            goal=goal,
+            plan=plan,
+            start_turn=current_turn,
+        )
+    
+    def advance_intention(self, success: bool) -> None:
+        """Record action execution and advance the intention."""
+        if not self.has_intention():
+            return
+        
+        self.current_intention.advance(success)
+        self.total_actions_executed += 1
+        
+        # Check if intention is complete
+        if self.current_intention.is_complete:
+            self.intentions_completed += 1
+            self.current_intention = None
+    
+    def abandon_intention(self, reason: str = "unknown") -> None:
+        """Abandon the current intention."""
+        if self.has_intention():
+            self.intentions_abandoned += 1
+            self.current_intention = None
+    
+    def get_plan_progress(self) -> dict:
+        """Get current plan execution progress."""
+        if not self.has_intention():
+            return {"has_intention": False}
+        
+        intention = self.current_intention
+        return {
+            "has_intention": True,
+            "goal": intention.goal.name,
+            "actions_completed": intention.actions_completed,
+            "remaining_actions": intention.remaining_actions,
+            "consecutive_failures": intention.consecutive_failures,
+            "turns_active": 0,  # Would need current_turn to calculate
+        }
+    
+    def to_dict(self) -> dict:
+        """Convert to dictionary for debugging/logging."""
+        return {
+            "commitment_strategy": self.commitment_strategy.value,
+            "has_intention": self.has_intention(),
+            "current_goal": self.current_intention.goal.name if self.has_intention() else None,
+            "plan_progress": self.get_plan_progress(),
+            "stats": {
+                "intentions_completed": self.intentions_completed,
+                "intentions_abandoned": self.intentions_abandoned,
+                "total_actions": self.total_actions_executed,
+            },
+        }

backend/core/engine.py

@@ -30,21 +30,26 @@ class TurnLog:
     turn: int
     agent_actions: list[dict] = field(default_factory=list)
     deaths: list[str] = field(default_factory=list)
+    births: list[str] = field(default_factory=list)
     trades: list[dict] = field(default_factory=list)
     # Resource tracking for this turn
     resources_produced: dict = field(default_factory=dict)
     resources_consumed: dict = field(default_factory=dict)
     resources_spoiled: dict = field(default_factory=dict)
+    # New day events
+    day_events: dict = field(default_factory=dict)
 
     def to_dict(self) -> dict:
         return {
             "turn": self.turn,
             "agent_actions": self.agent_actions,
             "deaths": self.deaths,
+            "births": self.births,
             "trades": self.trades,
             "resources_produced": self.resources_produced,
             "resources_consumed": self.resources_consumed,
             "resources_spoiled": self.resources_spoiled,
+            "day_events": self.day_events,
         }
 
 
@@ -171,21 +176,15 @@ class GameEngine:
                 money=agent.money,
             )
 
-            if self.world.is_night():
-                # Force sleep at night
-                decision = AIDecision(
-                    action=ActionType.SLEEP,
-                    reason="Night time: sleeping",
-                )
-            else:
-                # Pass time info so AI can prepare for night
-                decision = get_ai_decision(
-                    agent,
-                    self.market,
-                    step_in_day=self.world.step_in_day,
-                    day_steps=self.world.config.day_steps,
-                    current_turn=current_turn,
-                )
+            # GOAP AI handles night time automatically
+            decision = get_ai_decision(
+                agent,
+                self.market,
+                step_in_day=self.world.step_in_day,
+                day_steps=self.world.config.day_steps,
+                current_turn=current_turn,
+                is_night=self.world.is_night(),
+            )
 
             decisions.append((agent, decision))
 
@@ -283,28 +282,54 @@ class GameEngine:
         # End turn logging
         self.logger.end_turn()
 
-        # 8. Advance time
-        self.world.advance_time()
+        # 8. Advance time (returns True if new day started)
+        new_day = self.world.advance_time()
 
-        # 9. Check win/lose conditions (count only truly living agents, not corpses)
+        # 9. Process new day events (aging, births, sinks)
+        if new_day:
+            day_events = self._process_new_day(turn_log)
+            turn_log.day_events = day_events
+
+        # 10. Check win/lose conditions (count only truly living agents, not corpses)
         if len(self.world.get_living_agents()) == 0:
             self.is_running = False
             self.logger.close()
 
+        # Keep turn_logs bounded to prevent memory growth
+        max_logs = get_config().performance.max_turn_logs
         self.turn_logs.append(turn_log)
+        if len(self.turn_logs) > max_logs:
+            # Remove oldest logs, keep only recent ones
+            self.turn_logs = self.turn_logs[-max_logs:]
         return turn_log
 
     def _mark_dead_agents(self, current_turn: int) -> list[Agent]:
-        """Mark agents who just died as corpses. Returns list of newly dead agents."""
+        """Mark agents who just died as corpses. Returns list of newly dead agents.
+
+        Also processes inheritance - distributing wealth to children.
+        """
         newly_dead = []
         for agent in self.world.agents:
             if not agent.is_alive() and not agent.is_corpse():
-                # Agent just died this turn
-                cause = agent.stats.get_critical_stat() or "unknown"
+                # Determine cause of death
+                if agent.is_too_old():
+                    cause = "age"
+                else:
+                    cause = agent.stats.get_critical_stat() or "unknown"
+
+                # Process inheritance BEFORE marking dead (while inventory still accessible)
+                inheritance = self.world.process_inheritance(agent)
+                if inheritance.get("beneficiaries"):
+                    self.logger.log_event("inheritance", inheritance)
+
                 agent.mark_dead(current_turn, cause)
                 # Clear their action to show death state
                 agent.current_action.action_type = "dead"
                 agent.current_action.message = f"Died: {cause}"
+
+                # Record death statistics
+                self.world.record_death(agent, cause)
+
                 newly_dead.append(agent)
         return newly_dead
 
@@ -318,10 +343,150 @@ class GameEngine:
 
         for agent in to_remove:
             self.world.agents.remove(agent)
-            self.world.total_agents_died += 1
+            # Remove from index as well
+            self.world._agent_index.pop(agent.id, None)
+            # Note: death was already recorded in _mark_dead_agents
 
         return to_remove
 
+    def _process_new_day(self, turn_log: TurnLog) -> dict:
+        """Process all new-day events: aging, births, resource sinks.
+
+        Called when a new simulation day starts.
+        """
+        events = {
+            "day": self.world.current_day,
+            "births": [],
+            "age_deaths": [],
+            "taxes_collected": 0,
+            "storage_decay": {},
+            "random_events": [],
+        }
+
+        sinks_config = get_config().sinks
+        age_config = get_config().age
+
+        # 1. Age all living agents
+        for agent in self.world.get_living_agents():
+            agent.age_one_day()
+
+        # 2. Check for age-related deaths (after aging)
+        current_turn = self.world.current_turn
+        for agent in self.world.agents:
+            if not agent.is_corpse() and agent.is_too_old() and not agent.is_alive():
+                # Will be caught by _mark_dead_agents in the next turn
+                pass
+
+        # 3. Process potential births
+        for agent in list(self.world.get_living_agents()):  # Copy list since we modify it
+            if agent.can_give_birth(self.world.current_day):
+                child = self.world.spawn_child(agent)
+                if child:
+                    birth_info = {
+                        "parent_id": agent.id,
+                        "parent_name": agent.name,
+                        "child_id": child.id,
+                        "child_name": child.name,
+                    }
+                    events["births"].append(birth_info)
+                    turn_log.births.append(child.name)
+                    self.logger.log_event("birth", birth_info)
+
+        # 4. Apply daily money tax (wealth redistribution/removal)
+        if sinks_config.daily_tax_rate > 0:
+            total_taxes = 0
+            for agent in self.world.get_living_agents():
+                tax = int(agent.money * sinks_config.daily_tax_rate)
+                if tax > 0:
+                    agent.money -= tax
+                    total_taxes += tax
+            events["taxes_collected"] = total_taxes
+
+        # 5. Apply village storage decay (resources spoil over time)
+        if sinks_config.daily_village_decay_rate > 0:
+            decay_rate = sinks_config.daily_village_decay_rate
+            for agent in self.world.get_living_agents():
+                for resource in agent.inventory[:]:  # Copy list to allow modification
+                    # Random chance for each resource to decay
+                    if random.random() < decay_rate:
+                        decay_amount = max(1, int(resource.quantity * decay_rate))
+                        resource.quantity -= decay_amount
+
+                        res_type = resource.type.value
+                        events["storage_decay"][res_type] = events["storage_decay"].get(res_type, 0) + decay_amount
+
+                        # Track as spoiled
+                        turn_log.resources_spoiled[res_type] = turn_log.resources_spoiled.get(res_type, 0) + decay_amount
+                        self.resource_stats["spoiled"][res_type] = self.resource_stats["spoiled"].get(res_type, 0) + decay_amount
+
+                        if resource.quantity <= 0:
+                            agent.inventory.remove(resource)
+
+        # 6. Random events (fires, theft, etc.)
+        if random.random() < sinks_config.random_event_chance:
+            event = self._generate_random_event(events, turn_log)
+            if event:
+                events["random_events"].append(event)
+
+        return events
+
+    def _generate_random_event(self, events: dict, turn_log: TurnLog) -> Optional[dict]:
+        """Generate a random village event (disaster, theft, etc.)."""
+        sinks_config = get_config().sinks
+        living_agents = self.world.get_living_agents()
+
+        if not living_agents:
+            return None
+
+        event_types = ["fire", "theft", "blessing"]
+        event_type = random.choice(event_types)
+
+        event_info = {"type": event_type, "affected": []}
+
+        if event_type == "fire":
+            # Fire destroys some resources from random agents
+            num_affected = max(1, len(living_agents) // 5)  # 20% of agents affected
+            affected_agents = random.sample(living_agents, min(num_affected, len(living_agents)))
+
+            for agent in affected_agents:
+                for resource in agent.inventory[:]:
+                    loss = int(resource.quantity * sinks_config.fire_event_resource_loss)
+                    if loss > 0:
+                        resource.quantity -= loss
+                        res_type = resource.type.value
+                        turn_log.resources_spoiled[res_type] = turn_log.resources_spoiled.get(res_type, 0) + loss
+
+                        if resource.quantity <= 0:
+                            agent.inventory.remove(resource)
+
+                event_info["affected"].append(agent.name)
+
+        elif event_type == "theft":
+            # Some money is stolen from wealthy agents
+            wealthy_agents = [a for a in living_agents if a.money > 1000]
+            if wealthy_agents:
+                victim = random.choice(wealthy_agents)
+                stolen = int(victim.money * sinks_config.theft_event_money_loss)
+                victim.money -= stolen
+                event_info["affected"].append(victim.name)
+                event_info["amount_stolen"] = stolen
+
+        elif event_type == "blessing":
+            # Good harvest - some agents get bonus resources
+            lucky_agent = random.choice(living_agents)
+            from backend.domain.resources import Resource, ResourceType
+            bonus_type = random.choice([ResourceType.BERRIES, ResourceType.WOOD])
+            bonus = Resource(type=bonus_type, quantity=random.randint(2, 5), created_turn=self.world.current_turn)
+            lucky_agent.add_to_inventory(bonus)
+            event_info["affected"].append(lucky_agent.name)
+            event_info["bonus"] = f"+{bonus.quantity} {bonus_type.value}"
+
+        if event_info["affected"]:
+            self.logger.log_event("random_event", event_info)
+            return event_info
+
+        return None
+
     def _execute_action(self, agent: Agent, decision: AIDecision) -> Optional[ActionResult]:
         """Execute an action for an agent."""
         action = decision.action
@@ -400,9 +565,17 @@ class GameEngine:
         - Gathering skill affects gather output
         - Woodcutting skill affects wood output
         - Skills improve with use
+
+        Age modifies:
+        - Energy costs (young use less, old use more)
+        - Skill effectiveness (young less effective, old more effective "wisdom")
+        - Learning rate (young learn faster, old learn slower)
         """
-        # Check energy
-        energy_cost = abs(config.energy_cost)
+        # Calculate age-modified energy cost
+        base_energy_cost = abs(config.energy_cost)
+        energy_cost_modifier = agent.get_energy_cost_modifier()
+        energy_cost = max(1, int(base_energy_cost * energy_cost_modifier))
+
         if not agent.spend_energy(energy_cost):
             return ActionResult(
                 action_type=action,
@@ -430,16 +603,20 @@ class GameEngine:
         # Get relevant skill for this action
         skill_name = self._get_skill_for_action(action)
         skill_value = getattr(agent.skills, skill_name, 1.0) if skill_name else 1.0
-        skill_modifier = get_action_skill_modifier(skill_value)
 
-        # Check success chance (modified by skill)
+        # Apply age-based skill modifier (young less effective, old more effective)
+        age_skill_modifier = agent.get_skill_modifier()
+        skill_modifier = get_action_skill_modifier(skill_value) * age_skill_modifier
+
+        # Check success chance (modified by skill and age)
         # Higher skill = higher effective success chance
         effective_success_chance = min(0.98, config.success_chance * skill_modifier)
         if random.random() > effective_success_chance:
             # Record action attempt (skill still improves on failure, just less)
             agent.record_action(action.value)
             if skill_name:
-                agent.skills.improve(skill_name, 0.005)  # Small improvement on failure
+                learning_modifier = agent.get_learning_modifier()
+                agent.skills.improve(skill_name, 0.005, learning_modifier)  # Small improvement on failure
             return ActionResult(
                 action_type=action,
                 success=False,
@@ -447,14 +624,21 @@ class GameEngine:
                 message="Action failed",
             )
 
-        # Generate output (modified by skill for quantity)
+        # Generate output (modified by skill and age for quantity)
         resources_gained = []
 
         if config.output_resource:
+            # Check storage limit before producing
+            res_type = config.output_resource.value
+            storage_available = self.world.get_storage_available(res_type)
+
             # Skill affects output quantity
             base_quantity = random.randint(config.min_output, config.max_output)
             quantity = max(config.min_output, int(base_quantity * skill_modifier))
 
+            # Limit by storage
+            quantity = min(quantity, storage_available)
+
             if quantity > 0:
                 resource = Resource(
                     type=config.output_resource,
@@ -469,10 +653,15 @@ class GameEngine:
                         created_turn=self.world.current_turn,
                     ))
 
-        # Secondary output (e.g., hide from hunting) - also affected by skill
+        # Secondary output (e.g., hide from hunting) - also affected by skill and storage
         if config.secondary_output:
+            res_type = config.secondary_output.value
+            storage_available = self.world.get_storage_available(res_type)
+
             base_quantity = random.randint(config.secondary_min, config.secondary_max)
             quantity = max(0, int(base_quantity * skill_modifier))
+            quantity = min(quantity, storage_available)
+
             if quantity > 0:
                 resource = Resource(
                     type=config.secondary_output,
@@ -487,10 +676,11 @@ class GameEngine:
                         created_turn=self.world.current_turn,
                     ))
 
-        # Record action and improve skill
+        # Record action and improve skill (modified by age learning rate)
         agent.record_action(action.value)
         if skill_name:
-            agent.skills.improve(skill_name, 0.015)  # Skill improves with successful use
+            learning_modifier = agent.get_learning_modifier()
+            agent.skills.improve(skill_name, 0.015, learning_modifier)  # Skill improves with successful use
 
         # Build success message with details
         gained_str = ", ".join(f"+{r.quantity} {r.type.value}" for r in resources_gained)
@@ -571,13 +761,15 @@ class GameEngine:
                 if seller:
                     seller.money += result.total_paid
                     seller.record_trade(result.total_paid)
-                    seller.skills.improve("trading", 0.02)  # Seller skill improves
+                    seller.skills.improve("trading", 0.02, seller.get_learning_modifier())  # Seller skill improves
 
-                agent.spend_energy(abs(config.energy_cost))
+                # Age-modified energy cost for trading
+                energy_cost = max(1, int(abs(config.energy_cost) * agent.get_energy_cost_modifier()))
+                agent.spend_energy(energy_cost)
 
-                # Record buyer's trade and improve skill
+                # Record buyer's trade and improve skill (with age learning modifier)
                 agent.record_action("trade")
-                agent.skills.improve("trading", 0.01)  # Buyer skill improves less
+                agent.skills.improve("trading", 0.01, agent.get_learning_modifier())  # Buyer skill improves less
 
                 return ActionResult(
                     action_type=ActionType.TRADE,

backend/core/goap/__init__.py

@@ -0,0 +1,39 @@
+"""GOAP (Goal-Oriented Action Planning) module for agent decision making.
+
+This module provides a GOAP-based AI system where agents:
+1. Evaluate their current world state
+2. Select the most relevant goal based on priorities
+3. Plan a sequence of actions to achieve that goal
+4. Execute the first action in the plan
+
+Key components:
+- WorldState: Dictionary-like representation of agent/world state
+- Goal: Goals with dynamic priority calculation
+- GOAPAction: Actions with preconditions and effects
+- Planner: A* search for finding optimal action sequences
+"""
+
+from .world_state import WorldState
+from .goal import Goal, GoalType
+from .action import GOAPAction
+from .planner import GOAPPlanner
+from .goals import SURVIVAL_GOALS, ECONOMIC_GOALS, get_all_goals
+from .actions import get_all_actions, get_action_by_type
+from .debug import GOAPDebugInfo, get_goap_debug_info, get_all_agents_goap_debug
+
+__all__ = [
+    'WorldState',
+    'Goal',
+    'GoalType',
+    'GOAPAction',
+    'GOAPPlanner',
+    'SURVIVAL_GOALS',
+    'ECONOMIC_GOALS',
+    'get_all_goals',
+    'get_all_actions',
+    'get_action_by_type',
+    'GOAPDebugInfo',
+    'get_goap_debug_info',
+    'get_all_agents_goap_debug',
+]
+

backend/core/goap/action.py

@@ -0,0 +1,419 @@
+"""GOAP Action definitions.
+
+Actions are the building blocks of plans. Each action has:
+- Preconditions: What must be true for the action to be valid
+- Effects: How the action changes the world state
+- Cost: How expensive the action is (for planning)
+"""
+
+from dataclasses import dataclass, field
+from typing import Callable, Optional, TYPE_CHECKING
+
+from backend.domain.action import ActionType
+from backend.domain.resources import ResourceType
+
+from .world_state import WorldState
+
+if TYPE_CHECKING:
+    from backend.domain.agent import Agent
+    from backend.core.market import OrderBook
+
+
+@dataclass
+class GOAPAction:
+    """A GOAP action that can be part of a plan.
+
+    Actions transform the world state. The planner uses preconditions
+    and effects to search for valid action sequences.
+
+    Attributes:
+        name: Human-readable name
+        action_type: The underlying ActionType to execute
+        target_resource: Optional resource this action targets
+        preconditions: Function that checks if action is valid in a state
+        effects: Function that returns the expected effects on state
+        cost: Function that calculates action cost (lower = preferred)
+        get_decision_params: Function to get parameters for AIDecision
+    """
+    name: str
+    action_type: ActionType
+    target_resource: Optional[ResourceType] = None
+
+    # Functions that evaluate in context of world state
+    preconditions: Callable[[WorldState], bool] = field(default=lambda s: True)
+    effects: Callable[[WorldState], dict] = field(default=lambda s: {})
+    cost: Callable[[WorldState], float] = field(default=lambda s: 1.0)
+
+    # For generating the actual decision
+    get_decision_params: Optional[Callable[[WorldState, "Agent", "OrderBook"], dict]] = None
+
+    def is_valid(self, state: WorldState) -> bool:
+        """Check if this action can be performed in the given state."""
+        return self.preconditions(state)
+
+    def apply(self, state: WorldState) -> WorldState:
+        """Apply this action's effects to a state, returning a new state.
+
+        This is used by the planner for forward search.
+        """
+        new_state = state.copy()
+        effects = self.effects(state)
+
+        for key, value in effects.items():
+            if hasattr(new_state, key):
+                if isinstance(value, (int, float)):
+                    # For numeric values, handle both absolute and relative changes
+                    current = getattr(new_state, key)
+                    if isinstance(current, bool):
+                        setattr(new_state, key, bool(value))
+                    else:
+                        setattr(new_state, key, value)
+                else:
+                    setattr(new_state, key, value)
+
+        # Recalculate urgencies
+        new_state._calculate_urgencies()
+
+        return new_state
+
+    def get_cost(self, state: WorldState) -> float:
+        """Get the cost of this action in the given state."""
+        return self.cost(state)
+
+    def __repr__(self) -> str:
+        resource = f"({self.target_resource.value})" if self.target_resource else ""
+        return f"GOAPAction({self.name}{resource})"
+
+    def __hash__(self) -> int:
+        return hash((self.name, self.action_type, self.target_resource))
+
+    def __eq__(self, other) -> bool:
+        if not isinstance(other, GOAPAction):
+            return False
+        return (self.name == other.name and
+                self.action_type == other.action_type and
+                self.target_resource == other.target_resource)
+
+
+def create_consume_action(
+    resource_type: ResourceType,
+    stat_name: str,
+    stat_increase: float,
+    secondary_stat: Optional[str] = None,
+    secondary_increase: float = 0.0,
+) -> GOAPAction:
+    """Factory for creating consume resource actions."""
+    count_name = f"{resource_type.value}_count" if resource_type != ResourceType.BERRIES else "berries_count"
+    if resource_type == ResourceType.MEAT:
+        count_name = "meat_count"
+    elif resource_type == ResourceType.WATER:
+        count_name = "water_count"
+
+    # Map stat name to pct name
+    pct_name = f"{stat_name}_pct"
+    secondary_pct = f"{secondary_stat}_pct" if secondary_stat else None
+
+    def preconditions(state: WorldState) -> bool:
+        count = getattr(state, count_name, 0)
+        return count > 0
+
+    def effects(state: WorldState) -> dict:
+        result = {}
+        current = getattr(state, pct_name)
+        result[pct_name] = min(1.0, current + stat_increase)
+
+        if secondary_pct:
+            current_sec = getattr(state, secondary_pct)
+            result[secondary_pct] = min(1.0, current_sec + secondary_increase)
+
+        # Reduce resource count
+        current_count = getattr(state, count_name)
+        result[count_name] = max(0, current_count - 1)
+
+        # Update food count if consuming food
+        if resource_type in [ResourceType.MEAT, ResourceType.BERRIES]:
+            result["food_count"] = max(0, state.food_count - 1)
+
+        return result
+
+    def cost(state: WorldState) -> float:
+        # Consuming is very cheap - 0 energy cost
+        return 0.5
+
+    return GOAPAction(
+        name=f"Consume {resource_type.value}",
+        action_type=ActionType.CONSUME,
+        target_resource=resource_type,
+        preconditions=preconditions,
+        effects=effects,
+        cost=cost,
+    )
+
+
+def create_gather_action(
+    action_type: ActionType,
+    resource_type: ResourceType,
+    energy_cost: float,
+    expected_output: int,
+    success_chance: float = 1.0,
+) -> GOAPAction:
+    """Factory for creating resource gathering actions."""
+    count_name = f"{resource_type.value}_count"
+    if resource_type == ResourceType.BERRIES:
+        count_name = "berries_count"
+    elif resource_type == ResourceType.MEAT:
+        count_name = "meat_count"
+
+    def preconditions(state: WorldState) -> bool:
+        # Need enough energy and inventory space
+        energy_needed = abs(energy_cost) / 50.0  # Convert to percentage
+        return state.energy_pct >= energy_needed + 0.05 and state.inventory_space > 0
+
+    def effects(state: WorldState) -> dict:
+        result = {}
+
+        # Spend energy
+        energy_spent = abs(energy_cost) / 50.0
+        result["energy_pct"] = max(0, state.energy_pct - energy_spent)
+
+        # Gain resources (adjusted for success chance)
+        effective_output = int(expected_output * success_chance)
+        current = getattr(state, count_name)
+        result[count_name] = current + effective_output
+
+        # Update food count if gathering food
+        if resource_type in [ResourceType.MEAT, ResourceType.BERRIES]:
+            result["food_count"] = state.food_count + effective_output
+
+        # Update inventory space
+        result["inventory_space"] = max(0, state.inventory_space - effective_output)
+
+        return result
+
+    def cost(state: WorldState) -> float:
+        # Calculate cost based on efficiency (energy per unit of food)
+        food_per_action = expected_output * success_chance
+        if food_per_action > 0:
+            base_cost = abs(energy_cost) / food_per_action * 0.5
+        else:
+            base_cost = abs(energy_cost) / 5.0
+
+        # Adjust for success chance (penalize unreliable actions slightly)
+        if success_chance < 1.0:
+            base_cost *= 1.0 + (1.0 - success_chance) * 0.3
+
+        # STRONG profession specialization effect for gathering
+        if action_type == ActionType.GATHER:
+            # Compare gather_preference to other preferences
+            # Specialists get big discounts, generalists pay penalty
+            other_prefs = (state.hunt_preference + state.trade_preference) / 2
+            relative_strength = state.gather_preference / max(0.1, other_prefs)
+
+            # relative_strength > 1.0 means gathering is your specialty
+            # relative_strength < 1.0 means you're NOT a gatherer
+            if relative_strength >= 1.0:
+                # Specialist discount: up to 50% off
+                preference_modifier = 1.0 / relative_strength
+            else:
+                # Non-specialist penalty: up to 3x cost
+                preference_modifier = 1.0 + (1.0 - relative_strength) * 2.0
+
+            base_cost *= preference_modifier
+
+            # Skill reduces cost further (experienced = efficient)
+            # skill 0: no bonus, skill 1.0: 40% discount
+            skill_modifier = 1.0 - state.gathering_skill * 0.4
+            base_cost *= skill_modifier
+
+        return base_cost
+
+    return GOAPAction(
+        name=f"{action_type.value}",
+        action_type=action_type,
+        target_resource=resource_type,
+        preconditions=preconditions,
+        effects=effects,
+        cost=cost,
+    )
+
+
+def create_buy_action(resource_type: ResourceType) -> GOAPAction:
+    """Factory for creating market buy actions."""
+    can_buy_name = f"can_buy_{resource_type.value}"
+    if resource_type in [ResourceType.MEAT, ResourceType.BERRIES]:
+        can_buy_name = "can_buy_food"  # Simplified - we check specific later
+
+    count_name = f"{resource_type.value}_count"
+    if resource_type == ResourceType.BERRIES:
+        count_name = "berries_count"
+    elif resource_type == ResourceType.MEAT:
+        count_name = "meat_count"
+
+    price_name = f"{resource_type.value}_market_price"
+    if resource_type in [ResourceType.MEAT, ResourceType.BERRIES]:
+        price_name = "food_market_price"
+
+    def preconditions(state: WorldState) -> bool:
+        # Check specific availability
+        if resource_type == ResourceType.MEAT:
+            can_buy = state.can_buy_meat
+        elif resource_type == ResourceType.BERRIES:
+            can_buy = state.can_buy_berries
+        else:
+            can_buy = getattr(state, f"can_buy_{resource_type.value}", False)
+
+        return can_buy and state.inventory_space > 0
+
+    def effects(state: WorldState) -> dict:
+        result = {}
+
+        # Get price
+        if resource_type == ResourceType.MEAT:
+            price = state.food_market_price
+        elif resource_type == ResourceType.BERRIES:
+            price = state.food_market_price
+        else:
+            price = getattr(state, price_name, 10)
+
+        # Spend money
+        result["money"] = state.money - price
+
+        # Gain resource
+        current = getattr(state, count_name)
+        result[count_name] = current + 1
+
+        # Update food count if buying food
+        if resource_type in [ResourceType.MEAT, ResourceType.BERRIES]:
+            result["food_count"] = state.food_count + 1
+
+        # Spend small energy
+        result["energy_pct"] = max(0, state.energy_pct - 0.02)
+
+        # Update inventory
+        result["inventory_space"] = max(0, state.inventory_space - 1)
+
+        return result
+
+    def cost(state: WorldState) -> float:
+        # Trading cost is low (1 energy)
+        base_cost = 0.5
+
+        # MILD profession effect for trading (everyone should be able to trade)
+        # Traders get a bonus, but non-traders shouldn't be heavily penalized
+        # (trading benefits the whole economy)
+        other_prefs = (state.hunt_preference + state.gather_preference) / 2
+        relative_strength = state.trade_preference / max(0.1, other_prefs)
+
+        if relative_strength >= 1.0:
+            # Specialist discount: up to 40% off for dedicated traders
+            preference_modifier = max(0.6, 1.0 / relative_strength)
+        else:
+            # Mild non-specialist penalty: up to 50% cost increase
+            preference_modifier = 1.0 + (1.0 - relative_strength) * 0.5
+
+        base_cost *= preference_modifier
+
+        # Skill reduces cost (experienced traders are efficient)
+        # skill 0: no bonus, skill 1.0: 40% discount
+        skill_modifier = 1.0 - state.trading_skill * 0.4
+        base_cost *= skill_modifier
+
+        # Market affinity still has mild effect
+        base_cost *= (1.2 - state.market_affinity * 0.4)
+
+        # Check if it's a good deal
+        if resource_type == ResourceType.MEAT:
+            price = state.food_market_price
+        elif resource_type == ResourceType.BERRIES:
+            price = state.food_market_price
+        else:
+            price = getattr(state, price_name, 100)
+
+        # Higher price = higher cost (scaled for 100-500g price range)
+        # At fair value (~150g), multiplier is ~1.5x
+        # At min price (100g), multiplier is ~1.33x
+        base_cost *= (1.0 + price / 300.0)
+
+        return base_cost
+
+    return GOAPAction(
+        name=f"Buy {resource_type.value}",
+        action_type=ActionType.TRADE,
+        target_resource=resource_type,
+        preconditions=preconditions,
+        effects=effects,
+        cost=cost,
+    )
+
+
+def create_rest_action() -> GOAPAction:
+    """Create the rest action."""
+    def preconditions(state: WorldState) -> bool:
+        return state.energy_pct < 0.9  # Only rest if not full
+
+    def effects(state: WorldState) -> dict:
+        # Rest restores energy (12 out of 50 = 0.24)
+        return {
+            "energy_pct": min(1.0, state.energy_pct + 0.24),
+        }
+
+    def cost(state: WorldState) -> float:
+        # Resting is cheap but we prefer productive actions
+        return 2.0
+
+    return GOAPAction(
+        name="Rest",
+        action_type=ActionType.REST,
+        preconditions=preconditions,
+        effects=effects,
+        cost=cost,
+    )
+
+
+def create_build_fire_action() -> GOAPAction:
+    """Create the build fire action."""
+    def preconditions(state: WorldState) -> bool:
+        return state.wood_count > 0 and state.energy_pct >= 0.1
+
+    def effects(state: WorldState) -> dict:
+        return {
+            "heat_pct": min(1.0, state.heat_pct + 0.20),  # Fire gives 20 heat out of 100
+            "wood_count": max(0, state.wood_count - 1),
+            "energy_pct": max(0, state.energy_pct - 0.08),  # 4 energy cost
+        }
+
+    def cost(state: WorldState) -> float:
+        # Building fire is relatively cheap when we have wood
+        return 1.5
+
+    return GOAPAction(
+        name="Build Fire",
+        action_type=ActionType.BUILD_FIRE,
+        target_resource=ResourceType.WOOD,
+        preconditions=preconditions,
+        effects=effects,
+        cost=cost,
+    )
+
+
+def create_sleep_action() -> GOAPAction:
+    """Create the sleep action (for night)."""
+    def preconditions(state: WorldState) -> bool:
+        return state.is_night
+
+    def effects(state: WorldState) -> dict:
+        return {
+            "energy_pct": 1.0,  # Full energy restore
+        }
+
+    def cost(state: WorldState) -> float:
+        return 0.0  # Sleep is mandatory at night
+
+    return GOAPAction(
+        name="Sleep",
+        action_type=ActionType.SLEEP,
+        preconditions=preconditions,
+        effects=effects,
+        cost=cost,
+    )
+

backend/core/goap/actions.py

@@ -0,0 +1,399 @@
+"""Predefined GOAP actions for agents.
+
+Actions are organized by category:
+- Consume actions: Use resources from inventory
+- Gather actions: Produce resources
+- Trade actions: Buy/sell on market
+- Utility actions: Rest, sleep, build fire
+"""
+
+from typing import Optional, TYPE_CHECKING
+
+from backend.domain.action import ActionType, ACTION_CONFIG
+from backend.domain.resources import ResourceType
+
+from .world_state import WorldState
+from .action import (
+    GOAPAction,
+    create_consume_action,
+    create_gather_action,
+    create_buy_action,
+    create_rest_action,
+    create_build_fire_action,
+    create_sleep_action,
+)
+
+if TYPE_CHECKING:
+    from backend.domain.agent import Agent
+    from backend.core.market import OrderBook
+
+
+def _get_action_configs():
+    """Get action configurations from config."""
+    return ACTION_CONFIG
+
+
+# =============================================================================
+# CONSUME ACTIONS
+# =============================================================================
+
+def _create_drink_water() -> GOAPAction:
+    """Drink water to restore thirst."""
+    return create_consume_action(
+        resource_type=ResourceType.WATER,
+        stat_name="thirst",
+        stat_increase=0.50,  # 50 thirst out of 100
+    )
+
+
+def _create_eat_meat() -> GOAPAction:
+    """Eat meat to restore hunger (primary food source)."""
+    return create_consume_action(
+        resource_type=ResourceType.MEAT,
+        stat_name="hunger",
+        stat_increase=0.35,  # 35 hunger
+        secondary_stat="energy",
+        secondary_increase=0.24,  # 12 energy
+    )
+
+
+def _create_eat_berries() -> GOAPAction:
+    """Eat berries to restore hunger and some thirst."""
+    return create_consume_action(
+        resource_type=ResourceType.BERRIES,
+        stat_name="hunger",
+        stat_increase=0.10,  # 10 hunger
+        secondary_stat="thirst",
+        secondary_increase=0.04,  # 4 thirst
+    )
+
+
+CONSUME_ACTIONS = [
+    _create_drink_water(),
+    _create_eat_meat(),
+    _create_eat_berries(),
+]
+
+
+# =============================================================================
+# GATHER ACTIONS
+# =============================================================================
+
+def _create_get_water() -> GOAPAction:
+    """Get water from the river."""
+    config = _get_action_configs()[ActionType.GET_WATER]
+    return create_gather_action(
+        action_type=ActionType.GET_WATER,
+        resource_type=ResourceType.WATER,
+        energy_cost=config.energy_cost,
+        expected_output=1,
+        success_chance=1.0,
+    )
+
+
+def _create_gather_berries() -> GOAPAction:
+    """Gather berries (safe, reliable)."""
+    config = _get_action_configs()[ActionType.GATHER]
+    expected = (config.min_output + config.max_output) // 2
+    return create_gather_action(
+        action_type=ActionType.GATHER,
+        resource_type=ResourceType.BERRIES,
+        energy_cost=config.energy_cost,
+        expected_output=expected,
+        success_chance=1.0,
+    )
+
+
+def _create_hunt() -> GOAPAction:
+    """Hunt for meat (risky, high reward).
+
+    Hunt should be attractive because:
+    - Meat gives much more hunger than berries (35 vs 10)
+    - Meat also gives energy (12)
+    - You also get hide for clothes
+
+    Cost is balanced against gathering:
+    - Hunt: -7 energy, 70% success, 2-5 meat + 0-2 hide
+    - Gather: -3 energy, 100% success, 2-4 berries
+
+    Effective food per energy:
+    - Hunt: 3.5 meat avg * 0.7 = 2.45 meat = 2.45 * 35 hunger = 85.75 hunger for 7 energy = 12.25 hunger/energy
+    - Gather: 3 berries avg * 1.0 = 3 berries = 3 * 10 hunger = 30 hunger for 3 energy = 10 hunger/energy
+
+    So hunting is actually MORE efficient per energy for hunger! The cost should reflect this.
+    """
+    config = _get_action_configs()[ActionType.HUNT]
+    expected = (config.min_output + config.max_output) // 2
+
+    # Custom preconditions for hunting
+    def preconditions(state: WorldState) -> bool:
+        # Need more energy for hunting (but not excessively so)
+        energy_needed = abs(config.energy_cost) / 50.0 + 0.05
+        return state.energy_pct >= energy_needed and state.inventory_space >= 2
+
+    def effects(state: WorldState) -> dict:
+        # Account for success chance
+        effective_meat = int(expected * config.success_chance)
+        effective_hide = int(1 * config.success_chance)  # Average hide
+
+        energy_spent = abs(config.energy_cost) / 50.0
+
+        return {
+            "energy_pct": max(0, state.energy_pct - energy_spent),
+            "meat_count": state.meat_count + effective_meat,
+            "food_count": state.food_count + effective_meat,
+            "hide_count": state.hide_count + effective_hide,
+            "inventory_space": max(0, state.inventory_space - effective_meat - effective_hide),
+        }
+
+    def cost(state: WorldState) -> float:
+        # Hunt should be comparable to gather when considering value:
+        # - Hunt gives 3.5 meat avg (35 hunger each) = 122.5 hunger value
+        # - Gather gives 3 berries avg (10 hunger each) = 30 hunger value
+        # Hunt is 4x more valuable for hunger! So cost can be higher but not 4x.
+
+        # Base cost similar to gather
+        base_cost = 0.6
+
+        # Success chance penalty (small)
+        if config.success_chance < 1.0:
+            base_cost *= 1.0 + (1.0 - config.success_chance) * 0.2
+
+        # STRONG profession specialization effect for hunting
+        # Compare hunt_preference to other preferences
+        other_prefs = (state.gather_preference + state.trade_preference) / 2
+        relative_strength = state.hunt_preference / max(0.1, other_prefs)
+
+        # relative_strength > 1.0 means hunting is your specialty
+        if relative_strength >= 1.0:
+            # Specialist discount: up to 50% off
+            preference_modifier = 1.0 / relative_strength
+        else:
+            # Non-specialist penalty: up to 3x cost
+            preference_modifier = 1.0 + (1.0 - relative_strength) * 2.0
+
+        base_cost *= preference_modifier
+
+        # Skill reduces cost further (experienced hunters are efficient)
+        # skill 0: no bonus, skill 1.0: 40% discount
+        skill_modifier = 1.0 - state.hunting_skill * 0.4
+        base_cost *= skill_modifier
+
+        # Risk tolerance still has mild effect
+        risk_modifier = 1.0 + (0.5 - state.risk_tolerance) * 0.15
+        base_cost *= risk_modifier
+
+        # Big bonus if we have no meat - prioritize getting some
+        if state.meat_count == 0:
+            base_cost *= 0.6
+
+        # Bonus if low on food in general
+        if state.food_count < 2:
+            base_cost *= 0.8
+
+        return base_cost
+
+    return GOAPAction(
+        name="Hunt",
+        action_type=ActionType.HUNT,
+        target_resource=ResourceType.MEAT,
+        preconditions=preconditions,
+        effects=effects,
+        cost=cost,
+    )
+
+
+def _create_chop_wood() -> GOAPAction:
+    """Chop wood for fires."""
+    config = _get_action_configs()[ActionType.CHOP_WOOD]
+    expected = (config.min_output + config.max_output) // 2
+    return create_gather_action(
+        action_type=ActionType.CHOP_WOOD,
+        resource_type=ResourceType.WOOD,
+        energy_cost=config.energy_cost,
+        expected_output=expected,
+        success_chance=config.success_chance,
+    )
+
+
+def _create_weave_clothes() -> GOAPAction:
+    """Craft clothes from hide."""
+    config = _get_action_configs()[ActionType.WEAVE]
+
+    def preconditions(state: WorldState) -> bool:
+        return (
+            state.hide_count >= 1 and
+            not state.has_clothes and
+            state.energy_pct >= abs(config.energy_cost) / 50.0 + 0.05
+        )
+
+    def effects(state: WorldState) -> dict:
+        return {
+            "has_clothes": True,
+            "hide_count": state.hide_count - 1,
+            "energy_pct": max(0, state.energy_pct - abs(config.energy_cost) / 50.0),
+        }
+
+    def cost(state: WorldState) -> float:
+        return abs(config.energy_cost) / 3.0
+
+    return GOAPAction(
+        name="Weave Clothes",
+        action_type=ActionType.WEAVE,
+        target_resource=ResourceType.CLOTHES,
+        preconditions=preconditions,
+        effects=effects,
+        cost=cost,
+    )
+
+
+GATHER_ACTIONS = [
+    _create_get_water(),
+    _create_gather_berries(),
+    _create_hunt(),
+    _create_chop_wood(),
+    _create_weave_clothes(),
+]
+
+
+# =============================================================================
+# TRADE ACTIONS
+# =============================================================================
+
+def _create_buy_water() -> GOAPAction:
+    """Buy water from the market."""
+    return create_buy_action(ResourceType.WATER)
+
+
+def _create_buy_meat() -> GOAPAction:
+    """Buy meat from the market."""
+    return create_buy_action(ResourceType.MEAT)
+
+
+def _create_buy_berries() -> GOAPAction:
+    """Buy berries from the market."""
+    return create_buy_action(ResourceType.BERRIES)
+
+
+def _create_buy_wood() -> GOAPAction:
+    """Buy wood from the market."""
+    return create_buy_action(ResourceType.WOOD)
+
+
+def _create_sell_action(resource_type: ResourceType, min_keep: int = 1) -> GOAPAction:
+    """Factory for creating sell actions."""
+    count_name = f"{resource_type.value}_count"
+    if resource_type == ResourceType.BERRIES:
+        count_name = "berries_count"
+    elif resource_type == ResourceType.MEAT:
+        count_name = "meat_count"
+
+    def preconditions(state: WorldState) -> bool:
+        current = getattr(state, count_name)
+        return current > min_keep and state.energy_pct >= 0.05
+
+    def effects(state: WorldState) -> dict:
+        # Estimate we'll get a reasonable price (around min_price from config)
+        # This is approximate - actual execution will get real prices
+        estimated_price = 100  # Base estimate (min_price from config)
+
+        current = getattr(state, count_name)
+        sell_qty = min(3, current - min_keep)  # Sell up to 3, keep minimum
+
+        result = {
+            "money": state.money + estimated_price * sell_qty,
+            count_name: current - sell_qty,
+            "inventory_space": state.inventory_space + sell_qty,
+            "energy_pct": max(0, state.energy_pct - 0.02),
+        }
+
+        # Update food count if selling food
+        if resource_type in [ResourceType.MEAT, ResourceType.BERRIES]:
+            result["food_count"] = state.food_count - sell_qty
+
+        return result
+
+    def cost(state: WorldState) -> float:
+        # Selling has low cost - everyone should be able to sell excess
+        base_cost = 1.0
+
+        # MILD profession effect for selling (everyone should be able to trade)
+        other_prefs = (state.hunt_preference + state.gather_preference) / 2
+        relative_strength = state.trade_preference / max(0.1, other_prefs)
+
+        if relative_strength >= 1.0:
+            # Specialist discount: up to 40% off for dedicated traders
+            preference_modifier = max(0.6, 1.0 / relative_strength)
+        else:
+            # Mild non-specialist penalty: up to 50% cost increase
+            preference_modifier = 1.0 + (1.0 - relative_strength) * 0.5
+
+        base_cost *= preference_modifier
+
+        # Skill reduces cost (experienced traders know the market)
+        # skill 0: no bonus, skill 1.0: 40% discount
+        skill_modifier = 1.0 - state.trading_skill * 0.4
+        base_cost *= skill_modifier
+
+        # Hoarders reluctant to sell (mild effect)
+        base_cost *= (0.8 + state.hoarding_rate * 0.4)
+
+        return base_cost
+
+    return GOAPAction(
+        name=f"Sell {resource_type.value}",
+        action_type=ActionType.TRADE,
+        target_resource=resource_type,
+        preconditions=preconditions,
+        effects=effects,
+        cost=cost,
+    )
+
+
+TRADE_ACTIONS = [
+    _create_buy_water(),
+    _create_buy_meat(),
+    _create_buy_berries(),
+    _create_buy_wood(),
+    _create_sell_action(ResourceType.WATER, min_keep=2),
+    _create_sell_action(ResourceType.MEAT, min_keep=1),
+    _create_sell_action(ResourceType.BERRIES, min_keep=2),
+    _create_sell_action(ResourceType.WOOD, min_keep=1),
+    _create_sell_action(ResourceType.HIDE, min_keep=0),
+]
+
+
+# =============================================================================
+# UTILITY ACTIONS
+# =============================================================================
+
+UTILITY_ACTIONS = [
+    create_rest_action(),
+    create_build_fire_action(),
+    create_sleep_action(),
+]
+
+
+# =============================================================================
+# ALL ACTIONS
+# =============================================================================
+
+def get_all_actions() -> list[GOAPAction]:
+    """Get all available GOAP actions."""
+    return CONSUME_ACTIONS + GATHER_ACTIONS + TRADE_ACTIONS + UTILITY_ACTIONS
+
+
+def get_action_by_type(action_type: ActionType) -> list[GOAPAction]:
+    """Get all GOAP actions of a specific type."""
+    all_actions = get_all_actions()
+    return [a for a in all_actions if a.action_type == action_type]
+
+
+def get_action_by_name(name: str) -> Optional[GOAPAction]:
+    """Get a specific action by name."""
+    all_actions = get_all_actions()
+    for action in all_actions:
+        if action.name == name:
+            return action
+    return None
+

backend/core/goap/debug.py

@@ -0,0 +1,258 @@
+"""GOAP Debug utilities for visualization and analysis.
+
+Provides detailed information about GOAP decision-making for debugging
+and visualization purposes.
+"""
+
+from dataclasses import dataclass, field
+from typing import Optional, TYPE_CHECKING
+
+from .world_state import WorldState, create_world_state
+from .goal import Goal
+from .action import GOAPAction
+from .planner import GOAPPlanner, ReactivePlanner, Plan
+from .goals import get_all_goals
+from .actions import get_all_actions
+
+if TYPE_CHECKING:
+    from backend.domain.agent import Agent
+    from backend.core.market import OrderBook
+
+
+@dataclass
+class GoalDebugInfo:
+    """Debug information for a single goal."""
+    name: str
+    goal_type: str
+    priority: float
+    is_satisfied: bool
+    is_selected: bool = False
+
+    def to_dict(self) -> dict:
+        return {
+            "name": self.name,
+            "goal_type": self.goal_type,
+            "priority": round(self.priority, 2),
+            "is_satisfied": self.is_satisfied,
+            "is_selected": self.is_selected,
+        }
+
+
+@dataclass
+class ActionDebugInfo:
+    """Debug information for a single action."""
+    name: str
+    action_type: str
+    target_resource: Optional[str]
+    is_valid: bool
+    cost: float
+    is_in_plan: bool = False
+    plan_order: int = -1
+
+    def to_dict(self) -> dict:
+        return {
+            "name": self.name,
+            "action_type": self.action_type,
+            "target_resource": self.target_resource,
+            "is_valid": self.is_valid,
+            "cost": round(self.cost, 2),
+            "is_in_plan": self.is_in_plan,
+            "plan_order": self.plan_order,
+        }
+
+
+@dataclass
+class PlanDebugInfo:
+    """Debug information for the current plan."""
+    goal_name: str
+    actions: list[str]
+    total_cost: float
+    plan_length: int
+
+    def to_dict(self) -> dict:
+        return {
+            "goal_name": self.goal_name,
+            "actions": self.actions,
+            "total_cost": round(self.total_cost, 2),
+            "plan_length": self.plan_length,
+        }
+
+
+@dataclass
+class GOAPDebugInfo:
+    """Complete GOAP debug information for an agent."""
+    agent_id: str
+    agent_name: str
+    world_state: dict
+    goals: list[GoalDebugInfo]
+    actions: list[ActionDebugInfo]
+    current_plan: Optional[PlanDebugInfo]
+    selected_action: Optional[str]
+    decision_reason: str
+
+    def to_dict(self) -> dict:
+        return {
+            "agent_id": self.agent_id,
+            "agent_name": self.agent_name,
+            "world_state": self.world_state,
+            "goals": [g.to_dict() for g in self.goals],
+            "actions": [a.to_dict() for a in self.actions],
+            "current_plan": self.current_plan.to_dict() if self.current_plan else None,
+            "selected_action": self.selected_action,
+            "decision_reason": self.decision_reason,
+        }
+
+
+def get_goap_debug_info(
+    agent: "Agent",
+    market: "OrderBook",
+    step_in_day: int = 1,
+    day_steps: int = 10,
+    is_night: bool = False,
+) -> GOAPDebugInfo:
+    """Get detailed GOAP debug information for an agent.
+
+    This function performs the same planning as the actual AI,
+    but captures detailed information about the decision process.
+    """
+    # Create world state
+    state = create_world_state(
+        agent=agent,
+        market=market,
+        step_in_day=step_in_day,
+        day_steps=day_steps,
+        is_night=is_night,
+    )
+
+    # Get goals and actions
+    all_goals = get_all_goals()
+    all_actions = get_all_actions()
+
+    # Evaluate all goals
+    goal_infos = []
+    selected_goal = None
+    selected_plan = None
+
+    # Sort by priority
+    goals_with_priority = []
+    for goal in all_goals:
+        priority = goal.priority(state)
+        satisfied = goal.satisfied(state)
+        goals_with_priority.append((goal, priority, satisfied))
+
+    goals_with_priority.sort(key=lambda x: x[1], reverse=True)
+
+    # Try planning for each goal
+    planner = GOAPPlanner(max_iterations=50)
+
+    for goal, priority, satisfied in goals_with_priority:
+        if priority > 0 and not satisfied:
+            plan = planner.plan(state, goal, all_actions)
+            if plan and not plan.is_empty:
+                selected_goal = goal
+                selected_plan = plan
+                break
+
+    # Build goal debug info
+    for goal, priority, satisfied in goals_with_priority:
+        info = GoalDebugInfo(
+            name=goal.name,
+            goal_type=goal.goal_type.value,
+            priority=priority,
+            is_satisfied=satisfied,
+            is_selected=(goal == selected_goal),
+        )
+        goal_infos.append(info)
+
+    # Build action debug info
+    action_infos = []
+    plan_action_names = []
+    if selected_plan:
+        plan_action_names = [a.name for a in selected_plan.actions]
+
+    for action in all_actions:
+        is_valid = action.is_valid(state)
+        cost = action.get_cost(state) if is_valid else float('inf')
+
+        in_plan = action.name in plan_action_names
+        order = plan_action_names.index(action.name) if in_plan else -1
+
+        info = ActionDebugInfo(
+            name=action.name,
+            action_type=action.action_type.value,
+            target_resource=action.target_resource.value if action.target_resource else None,
+            is_valid=is_valid,
+            cost=cost if cost != float('inf') else -1,
+            is_in_plan=in_plan,
+            plan_order=order,
+        )
+        action_infos.append(info)
+
+    # Sort actions: plan actions first (by order), then valid actions, then invalid
+    action_infos.sort(key=lambda a: (
+        0 if a.is_in_plan else 1,
+        a.plan_order if a.is_in_plan else 999,
+        0 if a.is_valid else 1,
+        a.cost if a.cost >= 0 else 9999,
+    ))
+
+    # Build plan debug info
+    plan_info = None
+    if selected_plan:
+        plan_info = PlanDebugInfo(
+            goal_name=selected_plan.goal.name,
+            actions=[a.name for a in selected_plan.actions],
+            total_cost=selected_plan.total_cost,
+            plan_length=len(selected_plan.actions),
+        )
+
+    # Determine selected action and reason
+    selected_action = None
+    reason = "No plan found"
+
+    if is_night:
+        selected_action = "Sleep"
+        reason = "Night time: sleeping"
+    elif selected_plan and selected_plan.first_action:
+        selected_action = selected_plan.first_action.name
+        reason = f"{selected_plan.goal.name}: {selected_action}"
+    else:
+        # Fallback to reactive planning
+        reactive_planner = ReactivePlanner()
+        best_action = reactive_planner.select_best_action(state, all_goals, all_actions)
+        if best_action:
+            selected_action = best_action.name
+            reason = f"Reactive: {best_action.name}"
+
+            # Mark the reactive action in the action list
+            for action_info in action_infos:
+                if action_info.name == best_action.name:
+                    action_info.is_in_plan = True
+                    action_info.plan_order = 0
+
+    return GOAPDebugInfo(
+        agent_id=agent.id,
+        agent_name=agent.name,
+        world_state=state.to_dict(),
+        goals=goal_infos,
+        actions=action_infos,
+        current_plan=plan_info,
+        selected_action=selected_action,
+        decision_reason=reason,
+    )
+
+
+def get_all_agents_goap_debug(
+    agents: list["Agent"],
+    market: "OrderBook",
+    step_in_day: int = 1,
+    day_steps: int = 10,
+    is_night: bool = False,
+) -> list[GOAPDebugInfo]:
+    """Get GOAP debug info for all agents."""
+    return [
+        get_goap_debug_info(agent, market, step_in_day, day_steps, is_night)
+        for agent in agents
+        if agent.is_alive()
+    ]
+

backend/core/goap/goal.py

@@ -0,0 +1,185 @@
+"""Goal definitions for GOAP planning.
+
+Goals represent what an agent wants to achieve. Each goal has:
+- A name/type for identification
+- A condition that checks if the goal is satisfied
+- A priority function that determines how important the goal is
+- Optional target state values for the planner
+"""
+
+from dataclasses import dataclass, field
+from enum import Enum
+from typing import Callable, Optional
+
+from .world_state import WorldState
+
+
+class GoalType(Enum):
+    """Types of goals agents can pursue."""
+    # Survival goals - highest priority when needed
+    SATISFY_THIRST = "satisfy_thirst"
+    SATISFY_HUNGER = "satisfy_hunger"
+    MAINTAIN_HEAT = "maintain_heat"
+    RESTORE_ENERGY = "restore_energy"
+
+    # Resource goals - medium priority
+    STOCK_WATER = "stock_water"
+    STOCK_FOOD = "stock_food"
+    STOCK_WOOD = "stock_wood"
+    GET_CLOTHES = "get_clothes"
+
+    # Economic goals - lower priority but persistent
+    BUILD_WEALTH = "build_wealth"
+    SELL_EXCESS = "sell_excess"
+    FIND_DEALS = "find_deals"
+    TRADER_ARBITRAGE = "trader_arbitrage"
+
+    # Night behavior
+    SLEEP = "sleep"
+
+
+@dataclass
+class Goal:
+    """A goal that an agent can pursue.
+
+    Goals are the driving force of GOAP. The planner searches for
+    action sequences that transform the current world state into
+    one where the goal condition is satisfied.
+
+    Attributes:
+        goal_type: The type of goal (for identification)
+        name: Human-readable name
+        is_satisfied: Function that checks if goal is achieved in a state
+        get_priority: Function that calculates goal priority (higher = more important)
+        target_state: Optional dict of state values the goal aims for
+        max_plan_depth: Maximum actions to plan for this goal
+    """
+    goal_type: GoalType
+    name: str
+    is_satisfied: Callable[[WorldState], bool]
+    get_priority: Callable[[WorldState], float]
+    target_state: dict = field(default_factory=dict)
+    max_plan_depth: int = 3
+
+    def satisfied(self, state: WorldState) -> bool:
+        """Check if goal is satisfied in the given state."""
+        return self.is_satisfied(state)
+
+    def priority(self, state: WorldState) -> float:
+        """Get the priority of this goal in the given state."""
+        return self.get_priority(state)
+
+    def __repr__(self) -> str:
+        return f"Goal({self.name})"
+
+
+def create_survival_goal(
+    goal_type: GoalType,
+    name: str,
+    stat_name: str,
+    target_pct: float = 0.6,
+    base_priority: float = 10.0,
+) -> Goal:
+    """Factory for creating survival-related goals.
+
+    Survival goals have high priority when the relevant stat is low.
+    Priority scales with urgency.
+    """
+    urgency_name = f"{stat_name}_urgency"
+    pct_name = f"{stat_name}_pct"
+
+    def is_satisfied(state: WorldState) -> bool:
+        return getattr(state, pct_name) >= target_pct
+
+    def get_priority(state: WorldState) -> float:
+        urgency = getattr(state, urgency_name)
+        pct = getattr(state, pct_name)
+
+        if urgency <= 0:
+            return 0.0  # No need to pursue this goal
+
+        # Priority increases with urgency
+        # Critical urgency (>1.0) gives very high priority
+        priority = base_priority * urgency
+
+        # Extra boost when critical
+        if pct < state.critical_threshold:
+            priority *= 2.0
+
+        return priority
+
+    return Goal(
+        goal_type=goal_type,
+        name=name,
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+        target_state={pct_name: target_pct},
+        max_plan_depth=2,  # Survival should be quick
+    )
+
+
+def create_resource_stock_goal(
+    goal_type: GoalType,
+    name: str,
+    resource_name: str,
+    target_count: int,
+    base_priority: float = 5.0,
+) -> Goal:
+    """Factory for creating resource stockpiling goals.
+
+    Resource goals have moderate priority and aim to maintain reserves.
+    """
+    count_name = f"{resource_name}_count"
+
+    def is_satisfied(state: WorldState) -> bool:
+        return getattr(state, count_name) >= target_count
+
+    def get_priority(state: WorldState) -> float:
+        current = getattr(state, count_name)
+
+        if current >= target_count:
+            return 0.0  # Already have enough
+
+        # Priority based on how far from target
+        deficit = target_count - current
+        priority = base_priority * (deficit / target_count)
+
+        # Lower priority if survival is urgent
+        max_urgency = max(state.thirst_urgency, state.hunger_urgency, state.heat_urgency)
+        if max_urgency > 0.5:
+            priority *= 0.5
+
+        # Hoarders prioritize stockpiling more
+        priority *= (0.8 + state.hoarding_rate * 0.4)
+
+        # Evening boost - stock up before night
+        if state.is_evening:
+            priority *= 1.5
+
+        return priority
+
+    return Goal(
+        goal_type=goal_type,
+        name=name,
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+        target_state={count_name: target_count},
+        max_plan_depth=3,
+    )
+
+
+def create_economic_goal(
+    goal_type: GoalType,
+    name: str,
+    is_satisfied: Callable[[WorldState], bool],
+    get_priority: Callable[[WorldState], float],
+) -> Goal:
+    """Factory for creating economic/trading goals."""
+    return Goal(
+        goal_type=goal_type,
+        name=name,
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+        max_plan_depth=2,
+    )
+

backend/core/goap/goals.py

@@ -0,0 +1,411 @@
+"""Predefined goals for GOAP agents.
+
+Goals are organized into categories:
+- Survival goals: Immediate needs (thirst, hunger, heat, energy)
+- Resource goals: Building reserves
+- Economic goals: Trading and wealth building
+"""
+
+from .goal import Goal, GoalType, create_survival_goal, create_resource_stock_goal, create_economic_goal
+from .world_state import WorldState
+
+
+# =============================================================================
+# SURVIVAL GOALS
+# =============================================================================
+
+def _create_satisfy_thirst_goal() -> Goal:
+    """Satisfy immediate thirst."""
+    return create_survival_goal(
+        goal_type=GoalType.SATISFY_THIRST,
+        name="Satisfy Thirst",
+        stat_name="thirst",
+        target_pct=0.5,  # Want to get to 50%
+        base_priority=15.0,  # Highest base priority - thirst is most dangerous
+    )
+
+
+def _create_satisfy_hunger_goal() -> Goal:
+    """Satisfy immediate hunger."""
+    return create_survival_goal(
+        goal_type=GoalType.SATISFY_HUNGER,
+        name="Satisfy Hunger",
+        stat_name="hunger",
+        target_pct=0.5,
+        base_priority=12.0,
+    )
+
+
+def _create_maintain_heat_goal() -> Goal:
+    """Maintain body heat."""
+    return create_survival_goal(
+        goal_type=GoalType.MAINTAIN_HEAT,
+        name="Maintain Heat",
+        stat_name="heat",
+        target_pct=0.5,
+        base_priority=10.0,
+    )
+
+
+def _create_restore_energy_goal() -> Goal:
+    """Restore energy when low."""
+    def is_satisfied(state: WorldState) -> bool:
+        return state.energy_pct >= 0.4
+
+    def get_priority(state: WorldState) -> float:
+        if state.energy_pct >= 0.4:
+            return 0.0
+
+        # Priority increases as energy decreases
+        urgency = (0.4 - state.energy_pct) / 0.4
+
+        # But not if we have more urgent survival needs
+        max_vital_urgency = max(state.thirst_urgency, state.hunger_urgency, state.heat_urgency)
+        if max_vital_urgency > 1.5:
+            # Critical survival need - don't rest
+            return 0.0
+
+        base_priority = 6.0 * urgency
+
+        # Evening boost - want energy for night
+        if state.is_evening:
+            base_priority *= 1.5
+
+        return base_priority
+
+    return Goal(
+        goal_type=GoalType.RESTORE_ENERGY,
+        name="Restore Energy",
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+        target_state={"energy_pct": 0.6},
+        max_plan_depth=1,  # Just rest
+    )
+
+
+SURVIVAL_GOALS = [
+    _create_satisfy_thirst_goal(),
+    _create_satisfy_hunger_goal(),
+    _create_maintain_heat_goal(),
+    _create_restore_energy_goal(),
+]
+
+
+# =============================================================================
+# RESOURCE GOALS
+# =============================================================================
+
+def _create_stock_water_goal() -> Goal:
+    """Maintain water reserves."""
+    def is_satisfied(state: WorldState) -> bool:
+        target = int(2 * (0.5 + state.hoarding_rate))
+        return state.water_count >= target
+
+    def get_priority(state: WorldState) -> float:
+        target = int(2 * (0.5 + state.hoarding_rate))
+
+        if state.water_count >= target:
+            return 0.0
+
+        deficit = target - state.water_count
+        base_priority = 4.0 * (deficit / max(1, target))
+
+        # Lower if urgent survival needs
+        if max(state.thirst_urgency, state.hunger_urgency) > 1.0:
+            base_priority *= 0.3
+
+        # Evening boost
+        if state.is_evening:
+            base_priority *= 2.0
+
+        return base_priority
+
+    return Goal(
+        goal_type=GoalType.STOCK_WATER,
+        name="Stock Water",
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+        max_plan_depth=2,
+    )
+
+
+def _create_stock_food_goal() -> Goal:
+    """Maintain food reserves (meat + berries)."""
+    def is_satisfied(state: WorldState) -> bool:
+        target = int(3 * (0.5 + state.hoarding_rate))
+        return state.food_count >= target
+
+    def get_priority(state: WorldState) -> float:
+        target = int(3 * (0.5 + state.hoarding_rate))
+
+        if state.food_count >= target:
+            return 0.0
+
+        deficit = target - state.food_count
+        base_priority = 4.0 * (deficit / max(1, target))
+
+        # Lower if urgent survival needs
+        if max(state.thirst_urgency, state.hunger_urgency) > 1.0:
+            base_priority *= 0.3
+
+        # Evening boost
+        if state.is_evening:
+            base_priority *= 2.0
+
+        # Risk-takers may prefer hunting (more food per action)
+        base_priority *= (0.8 + state.risk_tolerance * 0.4)
+
+        return base_priority
+
+    return Goal(
+        goal_type=GoalType.STOCK_FOOD,
+        name="Stock Food",
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+        max_plan_depth=2,
+    )
+
+
+def _create_stock_wood_goal() -> Goal:
+    """Maintain wood reserves for fires."""
+    def is_satisfied(state: WorldState) -> bool:
+        target = int(2 * (0.5 + state.hoarding_rate))
+        return state.wood_count >= target
+
+    def get_priority(state: WorldState) -> float:
+        target = int(2 * (0.5 + state.hoarding_rate))
+
+        if state.wood_count >= target:
+            return 0.0
+
+        deficit = target - state.wood_count
+        base_priority = 3.0 * (deficit / max(1, target))
+
+        # Higher priority if heat is becoming an issue
+        if state.heat_urgency > 0.5:
+            base_priority *= 1.5
+
+        # Lower if urgent survival needs
+        if max(state.thirst_urgency, state.hunger_urgency) > 1.0:
+            base_priority *= 0.3
+
+        return base_priority
+
+    return Goal(
+        goal_type=GoalType.STOCK_WOOD,
+        name="Stock Wood",
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+        max_plan_depth=2,
+    )
+
+
+def _create_get_clothes_goal() -> Goal:
+    """Get clothes for heat protection."""
+    def is_satisfied(state: WorldState) -> bool:
+        return state.has_clothes
+
+    def get_priority(state: WorldState) -> float:
+        if state.has_clothes:
+            return 0.0
+
+        # Only pursue if we have hide
+        if state.hide_count < 1:
+            return 0.0
+
+        base_priority = 2.0
+
+        # Higher if heat is an issue
+        if state.heat_urgency > 0.3:
+            base_priority *= 1.5
+
+        return base_priority
+
+    return Goal(
+        goal_type=GoalType.GET_CLOTHES,
+        name="Get Clothes",
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+        max_plan_depth=1,
+    )
+
+
+RESOURCE_GOALS = [
+    _create_stock_water_goal(),
+    _create_stock_food_goal(),
+    _create_stock_wood_goal(),
+    _create_get_clothes_goal(),
+]
+
+
+# =============================================================================
+# ECONOMIC GOALS
+# =============================================================================
+
+def _create_build_wealth_goal() -> Goal:
+    """Accumulate money through trading."""
+    def is_satisfied(state: WorldState) -> bool:
+        return state.is_wealthy
+
+    def get_priority(state: WorldState) -> float:
+        if state.is_wealthy:
+            return 0.0
+
+        # Base priority scaled by wealth desire
+        base_priority = 2.0 * state.wealth_desire
+
+        # Only when survival is stable
+        max_urgency = max(state.thirst_urgency, state.hunger_urgency, state.heat_urgency)
+        if max_urgency > 0.5:
+            return 0.0
+
+        # Traders prioritize wealth more
+        if state.is_trader:
+            base_priority *= 2.0
+
+        return base_priority
+
+    return create_economic_goal(
+        goal_type=GoalType.BUILD_WEALTH,
+        name="Build Wealth",
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+    )
+
+
+def _create_sell_excess_goal() -> Goal:
+    """Sell excess resources on the market."""
+    def is_satisfied(state: WorldState) -> bool:
+        # Satisfied if inventory is not getting full
+        return state.inventory_space > 3
+
+    def get_priority(state: WorldState) -> float:
+        if state.inventory_space > 5:
+            return 0.0  # Plenty of space
+
+        # Priority increases as inventory fills
+        fullness = 1.0 - (state.inventory_space / 12.0)
+        base_priority = 3.0 * fullness
+
+        # Low hoarders sell more readily
+        base_priority *= (1.5 - state.hoarding_rate)
+
+        # Only when survival is stable
+        max_urgency = max(state.thirst_urgency, state.hunger_urgency)
+        if max_urgency > 0.5:
+            base_priority *= 0.5
+
+        return base_priority
+
+    return create_economic_goal(
+        goal_type=GoalType.SELL_EXCESS,
+        name="Sell Excess",
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+    )
+
+
+def _create_find_deals_goal() -> Goal:
+    """Find good deals on the market."""
+    def is_satisfied(state: WorldState) -> bool:
+        # This goal is never fully "satisfied" - it's opportunistic
+        return False
+
+    def get_priority(state: WorldState) -> float:
+        # Only pursue if we have money and market access
+        if state.money < 10:
+            return 0.0
+
+        # Check if there are deals available
+        has_deals = state.can_buy_water or state.can_buy_food or state.can_buy_wood
+        if not has_deals:
+            return 0.0
+
+        # Base priority from market affinity
+        base_priority = 2.0 * state.market_affinity
+
+        # Only when survival is stable
+        max_urgency = max(state.thirst_urgency, state.hunger_urgency)
+        if max_urgency > 0.5:
+            return 0.0
+
+        # Need inventory space
+        if state.inventory_space < 2:
+            return 0.0
+
+        return base_priority
+
+    return create_economic_goal(
+        goal_type=GoalType.FIND_DEALS,
+        name="Find Deals",
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+    )
+
+
+def _create_trader_arbitrage_goal() -> Goal:
+    """Trader-specific arbitrage goal (buy low, sell high)."""
+    def is_satisfied(state: WorldState) -> bool:
+        return False  # Always looking for opportunities
+
+    def get_priority(state: WorldState) -> float:
+        # Only for traders
+        if not state.is_trader:
+            return 0.0
+
+        # Need capital to trade
+        if state.money < 20:
+            return 1.0  # Low priority - need to sell something first
+
+        # Base priority for traders
+        base_priority = 5.0
+
+        # Only when survival is stable
+        max_urgency = max(state.thirst_urgency, state.hunger_urgency, state.heat_urgency)
+        if max_urgency > 0.3:
+            base_priority *= 0.5
+
+        return base_priority
+
+    return create_economic_goal(
+        goal_type=GoalType.TRADER_ARBITRAGE,
+        name="Trader Arbitrage",
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+    )
+
+
+def _create_sleep_goal() -> Goal:
+    """Sleep at night."""
+    def is_satisfied(state: WorldState) -> bool:
+        return not state.is_night  # Satisfied when it's not night
+
+    def get_priority(state: WorldState) -> float:
+        if not state.is_night:
+            return 0.0
+
+        # Highest priority at night
+        return 100.0
+
+    return Goal(
+        goal_type=GoalType.SLEEP,
+        name="Sleep",
+        is_satisfied=is_satisfied,
+        get_priority=get_priority,
+        max_plan_depth=1,
+    )
+
+
+ECONOMIC_GOALS = [
+    _create_build_wealth_goal(),
+    _create_sell_excess_goal(),
+    _create_find_deals_goal(),
+    _create_trader_arbitrage_goal(),
+    _create_sleep_goal(),
+]
+
+
+def get_all_goals() -> list[Goal]:
+    """Get all available goals."""
+    return SURVIVAL_GOALS + RESOURCE_GOALS + ECONOMIC_GOALS
+

backend/core/goap/goap_ai.py

@@ -0,0 +1,411 @@
+"""GOAP-based AI decision system for agents.
+
+This module provides the main interface for GOAP-based decision making
+using Goal-Oriented Action Planning.
+"""
+
+from dataclasses import dataclass, field
+from typing import Optional, TYPE_CHECKING
+
+from backend.domain.action import ActionType
+from backend.domain.resources import ResourceType
+from backend.domain.personality import get_trade_price_modifier
+
+from .world_state import WorldState, create_world_state
+from .goal import Goal
+from .action import GOAPAction
+from .planner import GOAPPlanner, ReactivePlanner, Plan
+from .goals import get_all_goals
+from .actions import get_all_actions
+
+if TYPE_CHECKING:
+    from backend.domain.agent import Agent
+    from backend.core.market import OrderBook
+    from backend.core.ai import AIDecision, TradeItem
+
+
+@dataclass
+class TradeItem:
+    """A single item to buy/sell in a trade."""
+    order_id: str
+    resource_type: ResourceType
+    quantity: int
+    price_per_unit: int
+
+
+@dataclass
+class AIDecision:
+    """A decision made by the AI for an agent."""
+    action: ActionType
+    target_resource: Optional[ResourceType] = None
+    order_id: Optional[str] = None
+    quantity: int = 1
+    price: int = 0
+    reason: str = ""
+    trade_items: list[TradeItem] = field(default_factory=list)
+    adjust_order_id: Optional[str] = None
+    new_price: Optional[int] = None
+
+    # GOAP-specific fields
+    goal_name: str = ""
+    plan_length: int = 0
+
+    def to_dict(self) -> dict:
+        return {
+            "action": self.action.value,
+            "target_resource": self.target_resource.value if self.target_resource else None,
+            "order_id": self.order_id,
+            "quantity": self.quantity,
+            "price": self.price,
+            "reason": self.reason,
+            "trade_items": [
+                {
+                    "order_id": t.order_id,
+                    "resource_type": t.resource_type.value,
+                    "quantity": t.quantity,
+                    "price_per_unit": t.price_per_unit,
+                }
+                for t in self.trade_items
+            ],
+            "adjust_order_id": self.adjust_order_id,
+            "new_price": self.new_price,
+            "goal_name": self.goal_name,
+            "plan_length": self.plan_length,
+        }
+
+
+class GOAPAgentAI:
+    """GOAP-based AI decision maker for agents.
+
+    This uses goal-oriented action planning to select actions:
+    1. Build world state from agent and market
+    2. Evaluate all goals and their priorities
+    3. Use planner to find action sequence for best goal
+    4. Return the first action as the decision
+
+    Falls back to reactive planning for simple decisions.
+    """
+
+    def __init__(
+        self,
+        agent: "Agent",
+        market: "OrderBook",
+        step_in_day: int = 1,
+        day_steps: int = 10,
+        current_turn: int = 0,
+        is_night: bool = False,
+    ):
+        self.agent = agent
+        self.market = market
+        self.step_in_day = step_in_day
+        self.day_steps = day_steps
+        self.current_turn = current_turn
+        self.is_night = is_night
+
+        # Build world state
+        self.state = create_world_state(
+            agent=agent,
+            market=market,
+            step_in_day=step_in_day,
+            day_steps=day_steps,
+            is_night=is_night,
+        )
+
+        # Initialize planners
+        self.planner = GOAPPlanner(max_iterations=50)
+        self.reactive_planner = ReactivePlanner()
+
+        # Get available goals and actions
+        self.goals = get_all_goals()
+        self.actions = get_all_actions()
+
+        # Personality shortcuts
+        self.p = agent.personality
+        self.skills = agent.skills
+
+    def decide(self) -> AIDecision:
+        """Make a decision using GOAP planning.
+
+        Decision flow:
+        1. Force sleep if night
+        2. Try to find a plan for the highest priority goal
+        3. If no plan found, use reactive selection
+        4. Convert GOAP action to AIDecision with proper parameters
+        """
+        # Night time - mandatory sleep
+        if self.is_night:
+            return AIDecision(
+                action=ActionType.SLEEP,
+                reason="Night time: sleeping",
+                goal_name="Sleep",
+            )
+
+        # Try GOAP planning
+        plan = self.planner.plan_for_goals(
+            initial_state=self.state,
+            goals=self.goals,
+            available_actions=self.actions,
+        )
+
+        if plan and not plan.is_empty:
+            # We have a plan - execute first action
+            goap_action = plan.first_action
+            return self._convert_to_decision(
+                goap_action=goap_action,
+                goal=plan.goal,
+                plan=plan,
+            )
+
+        # Fallback to reactive selection
+        best_action = self.reactive_planner.select_best_action(
+            state=self.state,
+            goals=self.goals,
+            available_actions=self.actions,
+        )
+
+        if best_action:
+            return self._convert_to_decision(
+                goap_action=best_action,
+                goal=None,
+                plan=None,
+            )
+
+        # Ultimate fallback - rest
+        return AIDecision(
+            action=ActionType.REST,
+            reason="No valid action found, resting",
+        )
+
+    def _convert_to_decision(
+        self,
+        goap_action: GOAPAction,
+        goal: Optional[Goal],
+        plan: Optional[Plan],
+    ) -> AIDecision:
+        """Convert a GOAP action to an AIDecision with proper parameters.
+
+        This handles the translation from abstract GOAP actions to
+        concrete decisions with order IDs, prices, etc.
+        """
+        action_type = goap_action.action_type
+        target_resource = goap_action.target_resource
+
+        # Build reason string
+        if goal:
+            reason = f"{goal.name}: {goap_action.name}"
+        else:
+            reason = f"Reactive: {goap_action.name}"
+
+        # Handle different action types
+        if action_type == ActionType.CONSUME:
+            return AIDecision(
+                action=action_type,
+                target_resource=target_resource,
+                reason=reason,
+                goal_name=goal.name if goal else "",
+                plan_length=len(plan.actions) if plan else 0,
+            )
+
+        elif action_type == ActionType.TRADE:
+            return self._create_trade_decision(goap_action, goal, plan, reason)
+
+        elif action_type in [ActionType.HUNT, ActionType.GATHER, ActionType.CHOP_WOOD,
+                            ActionType.GET_WATER, ActionType.WEAVE]:
+            return AIDecision(
+                action=action_type,
+                target_resource=target_resource,
+                reason=reason,
+                goal_name=goal.name if goal else "",
+                plan_length=len(plan.actions) if plan else 0,
+            )
+
+        elif action_type == ActionType.BUILD_FIRE:
+            return AIDecision(
+                action=action_type,
+                target_resource=ResourceType.WOOD,
+                reason=reason,
+                goal_name=goal.name if goal else "",
+                plan_length=len(plan.actions) if plan else 0,
+            )
+
+        elif action_type in [ActionType.REST, ActionType.SLEEP]:
+            return AIDecision(
+                action=action_type,
+                reason=reason,
+                goal_name=goal.name if goal else "",
+                plan_length=len(plan.actions) if plan else 0,
+            )
+
+        # Default case
+        return AIDecision(
+            action=action_type,
+            target_resource=target_resource,
+            reason=reason,
+            goal_name=goal.name if goal else "",
+            plan_length=len(plan.actions) if plan else 0,
+        )
+
+    def _create_trade_decision(
+        self,
+        goap_action: GOAPAction,
+        goal: Optional[Goal],
+        plan: Optional[Plan],
+        reason: str,
+    ) -> AIDecision:
+        """Create a trade decision with actual market parameters.
+
+        This translates abstract "Buy X" or "Sell X" actions into
+        concrete decisions with order IDs, prices, and quantities.
+        """
+        target_resource = goap_action.target_resource
+        action_name = goap_action.name.lower()
+
+        if "buy" in action_name:
+            # Find the best order to buy from
+            order = self.market.get_cheapest_order(target_resource)
+
+            if order and order.seller_id != self.agent.id:
+                # Calculate quantity to buy
+                can_afford = self.agent.money // max(1, order.price_per_unit)
+                space = self.agent.inventory_space()
+                quantity = min(2, can_afford, space, order.quantity)
+
+                if quantity > 0:
+                    return AIDecision(
+                        action=ActionType.TRADE,
+                        target_resource=target_resource,
+                        order_id=order.id,
+                        quantity=quantity,
+                        price=order.price_per_unit,
+                        reason=f"{reason} @ {order.price_per_unit}c",
+                        goal_name=goal.name if goal else "",
+                        plan_length=len(plan.actions) if plan else 0,
+                    )
+
+            # Can't buy - fallback to gathering
+            return self._create_gather_fallback(target_resource, reason, goal, plan)
+
+        elif "sell" in action_name:
+            # Create a sell order
+            quantity_available = self.agent.get_resource_count(target_resource)
+
+            # Calculate minimum to keep
+            min_keep = self._get_min_keep(target_resource)
+            quantity_to_sell = min(3, quantity_available - min_keep)
+
+            if quantity_to_sell > 0:
+                price = self._calculate_sell_price(target_resource)
+
+                return AIDecision(
+                    action=ActionType.TRADE,
+                    target_resource=target_resource,
+                    quantity=quantity_to_sell,
+                    price=price,
+                    reason=f"{reason} @ {price}c",
+                    goal_name=goal.name if goal else "",
+                    plan_length=len(plan.actions) if plan else 0,
+                )
+
+        # Invalid trade action - rest
+        return AIDecision(
+            action=ActionType.REST,
+            reason="Trade not possible",
+        )
+
+    def _create_gather_fallback(
+        self,
+        resource_type: ResourceType,
+        reason: str,
+        goal: Optional[Goal],
+        plan: Optional[Plan],
+    ) -> AIDecision:
+        """Create a gather action as fallback when buying isn't possible."""
+        # Map resource to gather action
+        action_map = {
+            ResourceType.WATER: ActionType.GET_WATER,
+            ResourceType.BERRIES: ActionType.GATHER,
+            ResourceType.MEAT: ActionType.HUNT,
+            ResourceType.WOOD: ActionType.CHOP_WOOD,
+        }
+
+        action = action_map.get(resource_type, ActionType.GATHER)
+
+        return AIDecision(
+            action=action,
+            target_resource=resource_type,
+            reason=f"{reason} (gathering instead)",
+            goal_name=goal.name if goal else "",
+            plan_length=len(plan.actions) if plan else 0,
+        )
+
+    def _get_min_keep(self, resource_type: ResourceType) -> int:
+        """Get minimum quantity to keep for survival."""
+        # Adjusted by hoarding rate
+        hoarding_mult = 0.5 + self.p.hoarding_rate
+
+        base_min = {
+            ResourceType.WATER: 2,
+            ResourceType.MEAT: 1,
+            ResourceType.BERRIES: 2,
+            ResourceType.WOOD: 1,
+            ResourceType.HIDE: 0,
+        }
+
+        return int(base_min.get(resource_type, 1) * hoarding_mult)
+
+    def _calculate_sell_price(self, resource_type: ResourceType) -> int:
+        """Calculate sell price based on fair value and market conditions."""
+        # Get energy cost to produce
+        from backend.core.ai import get_energy_cost
+        from backend.config import get_config
+
+        config = get_config()
+        economy = getattr(config, 'economy', None)
+        energy_to_money_ratio = getattr(economy, 'energy_to_money_ratio', 150) if economy else 150
+        min_price = getattr(economy, 'min_price', 100) if economy else 100
+
+        energy_cost = get_energy_cost(resource_type)
+        fair_value = max(min_price, int(energy_cost * energy_to_money_ratio))
+
+        # Apply trading skill
+        sell_modifier = get_trade_price_modifier(self.skills.trading, is_buying=False)
+
+        # Get market signal
+        signal = self.market.get_market_signal(resource_type)
+
+        if signal == "sell":  # Scarcity
+            price = int(fair_value * 1.3 * sell_modifier)
+        elif signal == "hold":
+            price = int(fair_value * sell_modifier)
+        else:  # Surplus
+            cheapest = self.market.get_cheapest_order(resource_type)
+            if cheapest and cheapest.seller_id != self.agent.id:
+                price = max(min_price, cheapest.price_per_unit - 1)
+            else:
+                price = int(fair_value * 0.8 * sell_modifier)
+
+        return max(min_price, price)
+
+
+def get_goap_decision(
+    agent: "Agent",
+    market: "OrderBook",
+    step_in_day: int = 1,
+    day_steps: int = 10,
+    current_turn: int = 0,
+    is_night: bool = False,
+) -> AIDecision:
+    """Convenience function to get a GOAP-based AI decision for an agent.
+
+    This is the main entry point for the GOAP AI system.
+    """
+    ai = GOAPAgentAI(
+        agent=agent,
+        market=market,
+        step_in_day=step_in_day,
+        day_steps=day_steps,
+        current_turn=current_turn,
+        is_night=is_night,
+    )
+    return ai.decide()
+

backend/core/goap/planner.py

@@ -0,0 +1,335 @@
+"""GOAP Planner using A* search.
+
+The planner finds optimal action sequences to achieve goals.
+It uses A* search with the goal condition as the target.
+"""
+
+import heapq
+from dataclasses import dataclass, field
+from typing import Optional
+
+from .world_state import WorldState
+from .goal import Goal
+from .action import GOAPAction
+
+
+@dataclass(order=True)
+class PlanNode:
+    """A node in the planning search tree."""
+    f_cost: float  # Total cost (g + h)
+    g_cost: float = field(compare=False)  # Cost so far
+    state: WorldState = field(compare=False)
+    action: Optional[GOAPAction] = field(compare=False, default=None)
+    parent: Optional["PlanNode"] = field(compare=False, default=None)
+    depth: int = field(compare=False, default=0)
+
+
+@dataclass
+class Plan:
+    """A plan is a sequence of actions to achieve a goal."""
+    goal: Goal
+    actions: list[GOAPAction]
+    total_cost: float
+    expected_final_state: WorldState
+
+    @property
+    def first_action(self) -> Optional[GOAPAction]:
+        """Get the first action to execute."""
+        return self.actions[0] if self.actions else None
+
+    @property
+    def is_empty(self) -> bool:
+        return len(self.actions) == 0
+
+    def __repr__(self) -> str:
+        action_names = " -> ".join(a.name for a in self.actions)
+        return f"Plan({self.goal.name}: {action_names} [cost={self.total_cost:.1f}])"
+
+
+class GOAPPlanner:
+    """A* planner for GOAP.
+
+    Finds the lowest-cost sequence of actions that transforms
+    the current world state into one where the goal is satisfied.
+    """
+
+    def __init__(self, max_iterations: int = 100):
+        self.max_iterations = max_iterations
+
+    def plan(
+        self,
+        initial_state: WorldState,
+        goal: Goal,
+        available_actions: list[GOAPAction],
+    ) -> Optional[Plan]:
+        """Find an action sequence to achieve the goal.
+
+        Uses A* search where:
+        - g(n) = accumulated action costs
+        - h(n) = heuristic estimate to goal (0 for now - effectively Dijkstra's)
+
+        Returns None if no plan found within iteration limit.
+        """
+        # Check if goal is already satisfied
+        if goal.satisfied(initial_state):
+            return Plan(
+                goal=goal,
+                actions=[],
+                total_cost=0.0,
+                expected_final_state=initial_state,
+            )
+
+        # Priority queue for A*
+        open_set: list[PlanNode] = []
+        start_node = PlanNode(
+            f_cost=0.0,
+            g_cost=0.0,
+            state=initial_state,
+            action=None,
+            parent=None,
+            depth=0,
+        )
+        heapq.heappush(open_set, start_node)
+
+        # Track visited states to avoid cycles
+        # We use a simplified state hash for efficiency
+        visited: set[tuple] = set()
+
+        iterations = 0
+
+        while open_set and iterations < self.max_iterations:
+            iterations += 1
+
+            # Get node with lowest f_cost
+            current = heapq.heappop(open_set)
+
+            # Check depth limit
+            if current.depth >= goal.max_plan_depth:
+                continue
+
+            # Create state hash for cycle detection
+            state_hash = self._hash_state(current.state)
+            if state_hash in visited:
+                continue
+            visited.add(state_hash)
+
+            # Try each action
+            for action in available_actions:
+                # Check if action is valid in current state
+                if not action.is_valid(current.state):
+                    continue
+
+                # Apply action to get new state
+                new_state = action.apply(current.state)
+
+                # Calculate costs
+                action_cost = action.get_cost(current.state)
+                g_cost = current.g_cost + action_cost
+                h_cost = self._heuristic(new_state, goal)
+                f_cost = g_cost + h_cost
+
+                # Create new node
+                new_node = PlanNode(
+                    f_cost=f_cost,
+                    g_cost=g_cost,
+                    state=new_state,
+                    action=action,
+                    parent=current,
+                    depth=current.depth + 1,
+                )
+
+                # Check if goal is satisfied
+                if goal.satisfied(new_state):
+                    # Reconstruct and return plan
+                    return self._reconstruct_plan(new_node, goal)
+
+                # Add to open set
+                heapq.heappush(open_set, new_node)
+
+        # No plan found
+        return None
+
+    def plan_for_goals(
+        self,
+        initial_state: WorldState,
+        goals: list[Goal],
+        available_actions: list[GOAPAction],
+    ) -> Optional[Plan]:
+        """Find the best plan among multiple goals.
+
+        Selects the highest-priority goal that has a valid plan,
+        considering both goal priority and plan cost.
+        """
+        # Sort goals by priority (highest first)
+        sorted_goals = sorted(goals, key=lambda g: g.priority(initial_state), reverse=True)
+
+        best_plan: Optional[Plan] = None
+        best_score = float('-inf')
+
+        for goal in sorted_goals:
+            priority = goal.priority(initial_state)
+
+            # Skip low-priority goals if we already have a good plan
+            if priority <= 0:
+                continue
+
+            if best_plan and priority < best_score * 0.5:
+                # This goal is much lower priority, skip
+                break
+
+            plan = self.plan(initial_state, goal, available_actions)
+
+            if plan:
+                # Score = priority / (cost + 1)
+                # Higher priority and lower cost = better
+                score = priority / (plan.total_cost + 1.0)
+
+                if score > best_score:
+                    best_score = score
+                    best_plan = plan
+
+        return best_plan
+
+    def _hash_state(self, state: WorldState) -> tuple:
+        """Create a hashable representation of key state values.
+
+        We don't hash everything - just the values that matter for planning.
+        """
+        return (
+            round(state.thirst_pct, 1),
+            round(state.hunger_pct, 1),
+            round(state.heat_pct, 1),
+            round(state.energy_pct, 1),
+            state.water_count,
+            state.food_count,
+            state.wood_count,
+            state.money // 10,  # Bucket money
+        )
+
+    def _heuristic(self, state: WorldState, goal: Goal) -> float:
+        """Estimate cost to reach goal from state.
+
+        For now, we use a simple heuristic based on the distance
+        from current state values to goal target values.
+        """
+        if not goal.target_state:
+            return 0.0
+
+        h = 0.0
+        for key, target in goal.target_state.items():
+            if hasattr(state, key):
+                current = getattr(state, key)
+                if isinstance(current, (int, float)) and isinstance(target, (int, float)):
+                    diff = abs(target - current)
+                    h += diff
+
+        return h
+
+    def _reconstruct_plan(self, final_node: PlanNode, goal: Goal) -> Plan:
+        """Reconstruct the action sequence from the final node."""
+        actions = []
+        node = final_node
+
+        while node.parent is not None:
+            if node.action:
+                actions.append(node.action)
+            node = node.parent
+
+        actions.reverse()
+
+        return Plan(
+            goal=goal,
+            actions=actions,
+            total_cost=final_node.g_cost,
+            expected_final_state=final_node.state,
+        )
+
+
+class ReactivePlanner:
+    """A simpler reactive planner for immediate needs.
+
+    Sometimes we don't need full planning - we just need to
+    pick the best immediate action. This planner evaluates
+    single actions against goals.
+    """
+
+    def select_best_action(
+        self,
+        state: WorldState,
+        goals: list[Goal],
+        available_actions: list[GOAPAction],
+    ) -> Optional[GOAPAction]:
+        """Select the single best action to take right now.
+
+        Evaluates each valid action and scores it based on how well
+        it progresses toward high-priority goals.
+        """
+        best_action: Optional[GOAPAction] = None
+        best_score = float('-inf')
+
+        for action in available_actions:
+            if not action.is_valid(state):
+                continue
+
+            score = self._score_action(state, action, goals)
+
+            if score > best_score:
+                best_score = score
+                best_action = action
+
+        return best_action
+
+    def _score_action(
+        self,
+        state: WorldState,
+        action: GOAPAction,
+        goals: list[Goal],
+    ) -> float:
+        """Score an action based on its contribution to goals."""
+        # Apply action to get expected new state
+        new_state = action.apply(state)
+        action_cost = action.get_cost(state)
+
+        total_score = 0.0
+
+        for goal in goals:
+            priority = goal.priority(state)
+            if priority <= 0:
+                continue
+
+            # Check if this action helps with the goal
+            was_satisfied = goal.satisfied(state)
+            now_satisfied = goal.satisfied(new_state)
+
+            if now_satisfied and not was_satisfied:
+                # Action satisfies the goal - big bonus!
+                total_score += priority * 10.0
+            elif not was_satisfied:
+                # Check if we made progress
+                # This is a simplified check based on urgencies
+                old_urgency = self._get_goal_urgency(goal, state)
+                new_urgency = self._get_goal_urgency(goal, new_state)
+
+                if new_urgency < old_urgency:
+                    improvement = old_urgency - new_urgency
+                    total_score += priority * improvement * 5.0
+
+        # Subtract cost
+        total_score -= action_cost
+
+        return total_score
+
+    def _get_goal_urgency(self, goal: Goal, state: WorldState) -> float:
+        """Get the urgency related to a goal."""
+        # Map goal types to state urgencies
+        from .goal import GoalType
+
+        urgency_map = {
+            GoalType.SATISFY_THIRST: state.thirst_urgency,
+            GoalType.SATISFY_HUNGER: state.hunger_urgency,
+            GoalType.MAINTAIN_HEAT: state.heat_urgency,
+            GoalType.RESTORE_ENERGY: state.energy_urgency,
+        }
+
+        return urgency_map.get(goal.goal_type, 0.0)
+

backend/core/goap/world_state.py

@@ -0,0 +1,319 @@
+"""World State representation for GOAP planning.
+
+The WorldState is a symbolic representation of the agent's current situation,
+used by the planner to reason about actions and goals.
+"""
+
+from dataclasses import dataclass, field
+from typing import TYPE_CHECKING, Optional
+
+if TYPE_CHECKING:
+    from backend.domain.agent import Agent
+    from backend.core.market import OrderBook
+
+
+@dataclass
+class WorldState:
+    """Symbolic representation of the world from an agent's perspective.
+
+    This captures all relevant state needed for GOAP planning:
+    - Agent vital stats (as percentages 0-1)
+    - Resource counts in inventory
+    - Market availability
+    - Economic state
+    - Time of day
+
+    The state uses normalized values (0-1) for stats to make
+    threshold comparisons easy and consistent.
+    """
+
+    # Vital stats as percentages (0.0 to 1.0)
+    thirst_pct: float = 1.0
+    hunger_pct: float = 1.0
+    heat_pct: float = 1.0
+    energy_pct: float = 1.0
+
+    # Resource counts in inventory
+    water_count: int = 0
+    food_count: int = 0  # meat + berries
+    meat_count: int = 0
+    berries_count: int = 0
+    wood_count: int = 0
+    hide_count: int = 0
+
+    # Inventory state
+    has_clothes: bool = False
+    inventory_space: int = 0
+    inventory_full: bool = False
+
+    # Economic state
+    money: int = 0
+    is_wealthy: bool = False  # Has comfortable money reserves
+
+    # Market availability (can we buy these?)
+    can_buy_water: bool = False
+    can_buy_food: bool = False
+    can_buy_meat: bool = False
+    can_buy_berries: bool = False
+    can_buy_wood: bool = False
+    water_market_price: int = 0
+    food_market_price: int = 0  # Cheapest of meat/berries
+    wood_market_price: int = 0
+
+    # Time state
+    is_night: bool = False
+    is_evening: bool = False  # Near end of day
+    step_in_day: int = 0
+    day_steps: int = 10
+
+    # Agent personality shortcuts (affect goal priorities)
+    wealth_desire: float = 0.5
+    hoarding_rate: float = 0.5
+    risk_tolerance: float = 0.5
+    market_affinity: float = 0.5
+    is_trader: bool = False
+
+    # Profession preferences (0.5-1.5 range, higher = more preferred)
+    gather_preference: float = 1.0
+    hunt_preference: float = 1.0
+    trade_preference: float = 1.0
+
+    # Skill levels (0.0-1.0, higher = more skilled)
+    hunting_skill: float = 0.0
+    gathering_skill: float = 0.0
+    trading_skill: float = 0.0
+
+    # Critical thresholds (from config)
+    critical_threshold: float = 0.25
+    low_threshold: float = 0.45
+
+    # Calculated urgencies (how urgent is each need?)
+    thirst_urgency: float = 0.0
+    hunger_urgency: float = 0.0
+    heat_urgency: float = 0.0
+    energy_urgency: float = 0.0
+
+    def __post_init__(self):
+        """Calculate urgencies after initialization."""
+        self._calculate_urgencies()
+
+    def _calculate_urgencies(self):
+        """Calculate urgency values for each vital stat.
+
+        Urgency is 0 when stat is full, and increases as stat decreases.
+        Urgency > 1.0 when in critical range.
+        """
+        # Urgency increases as stat decreases
+        # 0.0 = no urgency, 1.0 = needs attention, 2.0+ = critical
+
+        def calc_urgency(pct: float, critical: float, low: float) -> float:
+            if pct >= low:
+                return 0.0
+            elif pct >= critical:
+                # Linear increase from 0 to 1 as we go from low to critical
+                return 1.0 - (pct - critical) / (low - critical)
+            else:
+                # Exponential increase below critical
+                return 1.0 + (critical - pct) / critical * 2.0
+
+        self.thirst_urgency = calc_urgency(self.thirst_pct, self.critical_threshold, self.low_threshold)
+        self.hunger_urgency = calc_urgency(self.hunger_pct, self.critical_threshold, self.low_threshold)
+        self.heat_urgency = calc_urgency(self.heat_pct, self.critical_threshold, self.low_threshold)
+
+        # Energy urgency is different - we care about absolute level for work
+        if self.energy_pct < 0.25:
+            self.energy_urgency = 2.0
+        elif self.energy_pct < 0.40:
+            self.energy_urgency = 1.0
+        else:
+            self.energy_urgency = 0.0
+
+    def copy(self) -> "WorldState":
+        """Create a copy of this world state."""
+        return WorldState(
+            thirst_pct=self.thirst_pct,
+            hunger_pct=self.hunger_pct,
+            heat_pct=self.heat_pct,
+            energy_pct=self.energy_pct,
+            water_count=self.water_count,
+            food_count=self.food_count,
+            meat_count=self.meat_count,
+            berries_count=self.berries_count,
+            wood_count=self.wood_count,
+            hide_count=self.hide_count,
+            has_clothes=self.has_clothes,
+            inventory_space=self.inventory_space,
+            inventory_full=self.inventory_full,
+            money=self.money,
+            is_wealthy=self.is_wealthy,
+            can_buy_water=self.can_buy_water,
+            can_buy_food=self.can_buy_food,
+            can_buy_meat=self.can_buy_meat,
+            can_buy_berries=self.can_buy_berries,
+            can_buy_wood=self.can_buy_wood,
+            water_market_price=self.water_market_price,
+            food_market_price=self.food_market_price,
+            wood_market_price=self.wood_market_price,
+            is_night=self.is_night,
+            is_evening=self.is_evening,
+            step_in_day=self.step_in_day,
+            day_steps=self.day_steps,
+            wealth_desire=self.wealth_desire,
+            hoarding_rate=self.hoarding_rate,
+            risk_tolerance=self.risk_tolerance,
+            market_affinity=self.market_affinity,
+            is_trader=self.is_trader,
+            critical_threshold=self.critical_threshold,
+            low_threshold=self.low_threshold,
+        )
+
+    def to_dict(self) -> dict:
+        """Convert to dictionary for debugging/logging."""
+        return {
+            "vitals": {
+                "thirst": round(self.thirst_pct, 2),
+                "hunger": round(self.hunger_pct, 2),
+                "heat": round(self.heat_pct, 2),
+                "energy": round(self.energy_pct, 2),
+            },
+            "urgencies": {
+                "thirst": round(self.thirst_urgency, 2),
+                "hunger": round(self.hunger_urgency, 2),
+                "heat": round(self.heat_urgency, 2),
+                "energy": round(self.energy_urgency, 2),
+            },
+            "inventory": {
+                "water": self.water_count,
+                "meat": self.meat_count,
+                "berries": self.berries_count,
+                "wood": self.wood_count,
+                "hide": self.hide_count,
+                "space": self.inventory_space,
+            },
+            "economy": {
+                "money": self.money,
+                "is_wealthy": self.is_wealthy,
+            },
+            "market": {
+                "can_buy_water": self.can_buy_water,
+                "can_buy_food": self.can_buy_food,
+                "can_buy_wood": self.can_buy_wood,
+            },
+            "time": {
+                "is_night": self.is_night,
+                "is_evening": self.is_evening,
+                "step": self.step_in_day,
+            },
+        }
+
+
+def create_world_state(
+    agent: "Agent",
+    market: "OrderBook",
+    step_in_day: int = 1,
+    day_steps: int = 10,
+    is_night: bool = False,
+) -> WorldState:
+    """Create a WorldState from an agent and market.
+
+    This is the main factory function for creating world states.
+    It extracts all relevant information from the agent and market.
+    """
+    from backend.domain.resources import ResourceType
+    from backend.config import get_config
+
+    config = get_config()
+    agent_config = config.agent_stats
+    economy_config = getattr(config, 'economy', None)
+
+    stats = agent.stats
+
+    # Calculate stat percentages
+    thirst_pct = stats.thirst / stats.MAX_THIRST
+    hunger_pct = stats.hunger / stats.MAX_HUNGER
+    heat_pct = stats.heat / stats.MAX_HEAT
+    energy_pct = stats.energy / stats.MAX_ENERGY
+
+    # Get resource counts
+    water_count = agent.get_resource_count(ResourceType.WATER)
+    meat_count = agent.get_resource_count(ResourceType.MEAT)
+    berries_count = agent.get_resource_count(ResourceType.BERRIES)
+    wood_count = agent.get_resource_count(ResourceType.WOOD)
+    hide_count = agent.get_resource_count(ResourceType.HIDE)
+    food_count = meat_count + berries_count
+
+    # Check market availability
+    def get_market_info(resource_type: ResourceType) -> tuple[bool, int]:
+        """Get market availability and price for a resource."""
+        order = market.get_cheapest_order(resource_type)
+        if order and order.seller_id != agent.id and agent.money >= order.price_per_unit:
+            return True, order.price_per_unit
+        return False, 0
+
+    can_buy_water, water_price = get_market_info(ResourceType.WATER)
+    can_buy_meat, meat_price = get_market_info(ResourceType.MEAT)
+    can_buy_berries, berries_price = get_market_info(ResourceType.BERRIES)
+    can_buy_wood, wood_price = get_market_info(ResourceType.WOOD)
+
+    # Can buy food if we can buy either meat or berries
+    can_buy_food = can_buy_meat or can_buy_berries
+    food_price = min(
+        meat_price if can_buy_meat else float('inf'),
+        berries_price if can_buy_berries else float('inf')
+    )
+    food_price = food_price if food_price != float('inf') else 0
+
+    # Wealth calculation
+    min_wealth_target = getattr(economy_config, 'min_wealth_target', 50) if economy_config else 50
+    wealth_target = int(min_wealth_target * (0.5 + agent.personality.wealth_desire))
+    is_wealthy = agent.money >= wealth_target
+
+    # Trader check
+    is_trader = agent.personality.trade_preference > 1.3 and agent.personality.market_affinity > 0.5
+
+    # Evening check (last 2 steps before night)
+    is_evening = step_in_day >= day_steps - 2
+
+    return WorldState(
+        thirst_pct=thirst_pct,
+        hunger_pct=hunger_pct,
+        heat_pct=heat_pct,
+        energy_pct=energy_pct,
+        water_count=water_count,
+        food_count=food_count,
+        meat_count=meat_count,
+        berries_count=berries_count,
+        wood_count=wood_count,
+        hide_count=hide_count,
+        has_clothes=agent.has_clothes(),
+        inventory_space=agent.inventory_space(),
+        inventory_full=agent.inventory_full(),
+        money=agent.money,
+        is_wealthy=is_wealthy,
+        can_buy_water=can_buy_water,
+        can_buy_food=can_buy_food,
+        can_buy_meat=can_buy_meat,
+        can_buy_berries=can_buy_berries,
+        can_buy_wood=can_buy_wood,
+        water_market_price=water_price,
+        food_market_price=int(food_price),
+        wood_market_price=wood_price,
+        is_night=is_night,
+        is_evening=is_evening,
+        step_in_day=step_in_day,
+        day_steps=day_steps,
+        wealth_desire=agent.personality.wealth_desire,
+        hoarding_rate=agent.personality.hoarding_rate,
+        risk_tolerance=agent.personality.risk_tolerance,
+        market_affinity=agent.personality.market_affinity,
+        is_trader=is_trader,
+        gather_preference=agent.personality.gather_preference,
+        hunt_preference=agent.personality.hunt_preference,
+        trade_preference=agent.personality.trade_preference,
+        hunting_skill=agent.skills.hunting,
+        gathering_skill=agent.skills.gathering,
+        trading_skill=agent.skills.trading,
+        critical_threshold=agent_config.critical_threshold,
+        low_threshold=0.45,  # Could also be in config
+    )
+

backend/core/logger.py

@@ -1,7 +1,17 @@
-"""Simulation logger for detailed step-by-step logging."""
+"""Simulation logger for detailed step-by-step logging.
+
+Performance-optimized with non-blocking I/O:
+- Logging can be disabled or reduced via config
+- File writes happen in a background thread (producer-consumer pattern)
+- Agent lookups use O(1) dict instead of O(n) list search
+- No in-memory accumulation of all entries
+"""
 
 import json
 import logging
+import threading
+import queue
+import atexit
 from dataclasses import dataclass, field, asdict
 from datetime import datetime
 from pathlib import Path
@@ -39,7 +49,7 @@ class TurnLogEntry:
     trades: list = field(default_factory=list)
     deaths: list = field(default_factory=list)
     statistics: dict = field(default_factory=dict)
-    
+
     def to_dict(self) -> dict:
         """Convert to dictionary for JSON serialization."""
         return {
@@ -57,62 +67,250 @@ class TurnLogEntry:
         }
 
 
-class SimulationLogger:
-    """Logger that dumps detailed simulation data to files."""
+class AsyncLogWriter:
+    """Background thread that handles file I/O asynchronously.
     
+    Uses a producer-consumer pattern to decouple log generation
+    from file writing, preventing I/O from blocking the simulation.
+    """
+    
+    def __init__(self, max_queue_size: int = 1000):
+        self._queue: queue.Queue = queue.Queue(maxsize=max_queue_size)
+        self._stop_event = threading.Event()
+        self._thread: Optional[threading.Thread] = None
+        self._files: dict[str, TextIO] = {}
+        self._lock = threading.Lock()
+    
+    def start(self) -> None:
+        """Start the background writer thread."""
+        if self._thread is not None and self._thread.is_alive():
+            return
+        
+        self._stop_event.clear()
+        self._thread = threading.Thread(target=self._writer_loop, daemon=True)
+        self._thread.start()
+    
+    def stop(self, timeout: float = 2.0) -> None:
+        """Stop the background writer thread and flush remaining items."""
+        self._stop_event.set()
+        if self._thread is not None:
+            self._thread.join(timeout=timeout)
+            self._thread = None
+        
+        # Process any remaining items in the queue
+        self._drain_queue()
+        
+        # Close all files
+        with self._lock:
+            for f in self._files.values():
+                try:
+                    f.close()
+                except Exception:
+                    pass
+            self._files.clear()
+    
+    def _writer_loop(self) -> None:
+        """Main loop for the background writer thread."""
+        while not self._stop_event.is_set():
+            try:
+                # Wait for items with timeout to allow stop checks
+                item = self._queue.get(timeout=0.1)
+                self._process_item(item)
+                self._queue.task_done()
+            except queue.Empty:
+                continue
+            except Exception as e:
+                # Log errors but don't crash the thread
+                logging.getLogger("simulation").warning(f"Async log writer error: {e}")
+    
+    def _process_item(self, item: dict) -> None:
+        """Process a single log item."""
+        action = item.get("action")
+        
+        if action == "open":
+            self._open_file(item["path"], item["file_id"])
+        elif action == "write":
+            self._write_to_file(item["file_id"], item["data"])
+        elif action == "flush":
+            self._flush_file(item.get("file_id"))
+        elif action == "close":
+            self._close_file(item["file_id"])
+    
+    def _open_file(self, path: str, file_id: str) -> None:
+        """Open a file for writing."""
+        with self._lock:
+            if file_id not in self._files:
+                self._files[file_id] = open(path, "w", encoding="utf-8")
+    
+    def _write_to_file(self, file_id: str, data: str) -> None:
+        """Write data to a file."""
+        with self._lock:
+            f = self._files.get(file_id)
+            if f:
+                f.write(data)
+    
+    def _flush_file(self, file_id: Optional[str] = None) -> None:
+        """Flush file(s) to disk."""
+        with self._lock:
+            if file_id:
+                f = self._files.get(file_id)
+                if f:
+                    f.flush()
+            else:
+                for f in self._files.values():
+                    f.flush()
+    
+    def _close_file(self, file_id: str) -> None:
+        """Close a file."""
+        with self._lock:
+            f = self._files.pop(file_id, None)
+            if f:
+                f.close()
+    
+    def _drain_queue(self) -> None:
+        """Process all remaining items in the queue."""
+        while True:
+            try:
+                item = self._queue.get_nowait()
+                self._process_item(item)
+                self._queue.task_done()
+            except queue.Empty:
+                break
+    
+    def enqueue(self, item: dict) -> bool:
+        """Add an item to the write queue.
+        
+        Returns False if queue is full (item dropped).
+        """
+        try:
+            self._queue.put_nowait(item)
+            return True
+        except queue.Full:
+            return False
+    
+    def open_file(self, path: str, file_id: str) -> None:
+        """Queue a file open operation."""
+        self.enqueue({"action": "open", "path": path, "file_id": file_id})
+    
+    def write(self, file_id: str, data: str) -> None:
+        """Queue a write operation."""
+        self.enqueue({"action": "write", "file_id": file_id, "data": data})
+    
+    def flush(self, file_id: Optional[str] = None) -> None:
+        """Queue a flush operation."""
+        self.enqueue({"action": "flush", "file_id": file_id})
+    
+    def close_file(self, file_id: str) -> None:
+        """Queue a file close operation."""
+        self.enqueue({"action": "close", "file_id": file_id})
+
+
+class SimulationLogger:
+    """Logger that dumps detailed simulation data to files.
+
+    Performance optimized:
+    - Logging can be disabled entirely via config
+    - File writes happen in background thread (non-blocking)
+    - Agent lookups use O(1) dict instead of O(n) list search
+    - No in-memory accumulation of all entries
+    """
+
     def __init__(self, log_dir: str = "logs"):
         self.log_dir = Path(log_dir)
-        self.log_dir.mkdir(exist_ok=True)
+
+        # Load performance config
+        from backend.config import get_config
+        perf_config = get_config().performance
+        self.logging_enabled = perf_config.logging_enabled
+        self.detailed_logging = perf_config.detailed_logging
+        self.flush_interval = perf_config.log_flush_interval
         
-        # Create session-specific log file
-        timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
-        self.session_file = self.log_dir / f"sim_{timestamp}.jsonl"
-        self.summary_file = self.log_dir / f"sim_{timestamp}_summary.txt"
+        # Get async logging config (default to True if available)
+        self.async_logging = getattr(perf_config, 'async_logging', True)
+
+        # Async writer (only created if logging enabled)
+        self._async_writer: Optional[AsyncLogWriter] = None
         
-        # File handles
+        # File IDs for async writing
+        self._json_file_id: Optional[str] = None
+        self._summary_file_id: Optional[str] = None
+        
+        # Fallback: synchronous file handles (used if async disabled)
         self._json_file: Optional[TextIO] = None
         self._summary_file: Optional[TextIO] = None
-        
-        # Also set up standard Python logging
+
+        # Standard Python logging (minimal overhead even when enabled)
         self.logger = logging.getLogger("simulation")
-        self.logger.setLevel(logging.DEBUG)
-        
-        # File handler for detailed logs
-        file_handler = logging.FileHandler(self.log_dir / f"sim_{timestamp}.log")
-        file_handler.setLevel(logging.DEBUG)
-        file_handler.setFormatter(logging.Formatter(
-            "%(asctime)s | %(levelname)s | %(message)s"
-        ))
-        self.logger.addHandler(file_handler)
-        
-        # Console handler for important events
-        console_handler = logging.StreamHandler()
-        console_handler.setLevel(logging.INFO)
-        console_handler.setFormatter(logging.Formatter(
-            "%(asctime)s | %(message)s", datefmt="%H:%M:%S"
-        ))
-        self.logger.addHandler(console_handler)
-        
-        self._entries: list[TurnLogEntry] = []
+        self.logger.setLevel(logging.WARNING)  # Only warnings by default
+
+        # Current turn tracking
         self._current_entry: Optional[TurnLogEntry] = None
-    
+        # O(1) lookup for agent entries by ID
+        self._agent_entry_map: dict[str, AgentLogEntry] = {}
+
+        # Turn counter for flush batching
+        self._turns_since_flush = 0
+        
+        # Stats
+        self._items_queued = 0
+        self._items_dropped = 0
+
     def start_session(self, config: dict) -> None:
         """Start a new logging session."""
-        self._json_file = open(self.session_file, "w")
-        self._summary_file = open(self.summary_file, "w")
+        if not self.logging_enabled:
+            return
+
+        self.log_dir.mkdir(exist_ok=True)
+
+        # Create session-specific log file paths
+        timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+        json_path = self.log_dir / f"sim_{timestamp}.jsonl"
+        summary_path = self.log_dir / f"sim_{timestamp}_summary.txt"
         
-        # Write config as first line
-        self._json_file.write(json.dumps({"type": "config", "data": config}) + "\n")
-        self._json_file.flush()
-        
-        self._summary_file.write(f"Simulation Session Started: {datetime.now()}\n")
-        self._summary_file.write("=" * 60 + "\n\n")
-        self._summary_file.flush()
-        
-        self.logger.info(f"Logging session started: {self.session_file}")
-    
+        if self.async_logging:
+            # Use async writer
+            self._async_writer = AsyncLogWriter()
+            self._async_writer.start()
+            
+            self._json_file_id = f"json_{timestamp}"
+            self._summary_file_id = f"summary_{timestamp}"
+            
+            self._async_writer.open_file(str(json_path), self._json_file_id)
+            self._async_writer.open_file(str(summary_path), self._summary_file_id)
+            
+            # Write initial data
+            self._async_writer.write(
+                self._json_file_id,
+                json.dumps({"type": "config", "data": config}) + "\n"
+            )
+            self._async_writer.write(
+                self._summary_file_id,
+                f"Simulation Session Started: {datetime.now()}\n" + "=" * 60 + "\n\n"
+            )
+        else:
+            # Use synchronous file handles
+            self._json_file = open(json_path, "w")
+            self._summary_file = open(summary_path, "w")
+            
+            self._json_file.write(json.dumps({"type": "config", "data": config}) + "\n")
+            self._summary_file.write(f"Simulation Session Started: {datetime.now()}\n")
+            self._summary_file.write("=" * 60 + "\n\n")
+
+        if self.detailed_logging:
+            # Set up file handler for detailed logs
+            file_handler = logging.FileHandler(self.log_dir / f"sim_{timestamp}.log")
+            file_handler.setLevel(logging.DEBUG)
+            file_handler.setFormatter(logging.Formatter(
+                "%(asctime)s | %(levelname)s | %(message)s"
+            ))
+            self.logger.addHandler(file_handler)
+            self.logger.setLevel(logging.DEBUG)
+
     def start_turn(self, turn: int, day: int, step_in_day: int, time_of_day: str) -> None:
         """Start logging a new turn."""
+        if not self.logging_enabled:
+            return
+
         self._current_entry = TurnLogEntry(
             turn=turn,
             day=day,
@@ -120,8 +318,11 @@ class SimulationLogger:
             time_of_day=time_of_day,
             timestamp=datetime.now().isoformat(),
         )
-        self.logger.debug(f"Turn {turn} started (Day {day}, Step {step_in_day}, {time_of_day})")
-    
+        self._agent_entry_map.clear()
+
+        if self.detailed_logging:
+            self.logger.debug(f"Turn {turn} started (Day {day}, Step {step_in_day}, {time_of_day})")
+
     def log_agent_before(
         self,
         agent_id: str,
@@ -133,40 +334,42 @@ class SimulationLogger:
         money: int,
     ) -> None:
         """Log agent state before action."""
-        if self._current_entry is None:
+        if not self.logging_enabled or self._current_entry is None:
             return
-        
-        # Create placeholder entry
+
+        # Create entry and add to both list and map
         entry = AgentLogEntry(
             agent_id=agent_id,
             agent_name=agent_name,
             profession=profession,
-            position=position.copy(),
-            stats_before=stats.copy(),
+            position=position,
+            stats_before=stats,
             stats_after={},
             decision={},
             action_result={},
-            inventory_before=inventory.copy(),
+            inventory_before=inventory,
             inventory_after=[],
             money_before=money,
             money_after=money,
         )
         self._current_entry.agent_entries.append(entry)
-    
+        self._agent_entry_map[agent_id] = entry
+
     def log_agent_decision(self, agent_id: str, decision: dict) -> None:
         """Log agent's AI decision."""
-        if self._current_entry is None:
+        if not self.logging_enabled or self._current_entry is None:
             return
-        
-        for entry in self._current_entry.agent_entries:
-            if entry.agent_id == agent_id:
-                entry.decision = decision.copy()
+
+        # O(1) lookup instead of O(n) search
+        entry = self._agent_entry_map.get(agent_id)
+        if entry:
+            entry.decision = decision
+            if self.detailed_logging:
                 self.logger.debug(
                     f"  {entry.agent_name}: decided to {decision.get('action', '?')} "
                     f"- {decision.get('reason', '')}"
                 )
-                break
-    
+
     def log_agent_after(
         self,
         agent_id: str,
@@ -177,101 +380,169 @@ class SimulationLogger:
         action_result: dict,
     ) -> None:
         """Log agent state after action."""
-        if self._current_entry is None:
+        if not self.logging_enabled or self._current_entry is None:
             return
-        
-        for entry in self._current_entry.agent_entries:
-            if entry.agent_id == agent_id:
-                entry.stats_after = stats.copy()
-                entry.inventory_after = inventory.copy()
-                entry.money_after = money
-                entry.position = position.copy()
-                entry.action_result = action_result.copy()
-                break
-    
+
+        # O(1) lookup instead of O(n) search
+        entry = self._agent_entry_map.get(agent_id)
+        if entry:
+            entry.stats_after = stats
+            entry.inventory_after = inventory
+            entry.money_after = money
+            entry.position = position
+            entry.action_result = action_result
+
     def log_market_state(self, orders_before: list, orders_after: list) -> None:
         """Log market state."""
-        if self._current_entry is None:
+        if not self.logging_enabled or self._current_entry is None:
             return
         self._current_entry.market_orders_before = orders_before
         self._current_entry.market_orders_after = orders_after
-    
+
     def log_trade(self, trade: dict) -> None:
         """Log a trade transaction."""
-        if self._current_entry is None:
+        if not self.logging_enabled or self._current_entry is None:
             return
         self._current_entry.trades.append(trade)
-        self.logger.debug(f"  Trade: {trade.get('message', 'Unknown trade')}")
-    
+        if self.detailed_logging:
+            self.logger.debug(f"  Trade: {trade.get('message', 'Unknown trade')}")
+
     def log_death(self, agent_name: str, cause: str) -> None:
         """Log an agent death."""
-        if self._current_entry is None:
+        if not self.logging_enabled or self._current_entry is None:
             return
         self._current_entry.deaths.append({"name": agent_name, "cause": cause})
+        # Always log deaths even without detailed logging
         self.logger.info(f"  DEATH: {agent_name} died from {cause}")
-    
+
+    def log_event(self, event_type: str, event_data: dict) -> None:
+        """Log a general event (births, random events, etc.)."""
+        if not self.logging_enabled or self._current_entry is None:
+            return
+
+        if event_type == "birth":
+            self.logger.info(
+                f"  BIRTH: {event_data.get('child_name', '?')} born to {event_data.get('parent_name', '?')}"
+            )
+        elif event_type == "random_event" and self.detailed_logging:
+            self.logger.info(
+                f"  EVENT: {event_data.get('type', '?')} affecting {event_data.get('affected', [])}"
+            )
+        elif self.detailed_logging:
+            self.logger.debug(f"  Event [{event_type}]: {event_data}")
+
     def log_statistics(self, stats: dict) -> None:
         """Log end-of-turn statistics."""
-        if self._current_entry is None:
+        if not self.logging_enabled or self._current_entry is None:
             return
-        self._current_entry.statistics = stats.copy()
-    
+        self._current_entry.statistics = stats
+
     def end_turn(self) -> None:
         """Finish logging the current turn and write to file."""
-        if self._current_entry is None:
+        if not self.logging_enabled or self._current_entry is None:
             return
+
+        entry = self._current_entry
         
-        self._entries.append(self._current_entry)
+        # Prepare data
+        json_data = json.dumps({"type": "turn", "data": entry.to_dict()}) + "\n"
         
-        # Write to JSON lines file
-        if self._json_file:
-            self._json_file.write(
-                json.dumps({"type": "turn", "data": self._current_entry.to_dict()}) + "\n"
-            )
-            self._json_file.flush()
+        summary_lines = [f"Turn {entry.turn} | Day {entry.day} Step {entry.step_in_day} ({entry.time_of_day})\n"]
         
-        # Write summary
-        if self._summary_file:
-            entry = self._current_entry
-            self._summary_file.write(
-                f"Turn {entry.turn} | Day {entry.day} Step {entry.step_in_day} ({entry.time_of_day})\n"
-            )
-            
+        if self.detailed_logging:
             for agent in entry.agent_entries:
                 action = agent.decision.get("action", "?")
-                result = "✓" if agent.action_result.get("success", False) else "✗"
-                self._summary_file.write(
+                result = "+" if agent.action_result.get("success", False) else "-"
+                summary_lines.append(
                     f"  [{agent.agent_name}] {action} {result} | "
                     f"E:{agent.stats_after.get('energy', '?')} "
                     f"H:{agent.stats_after.get('hunger', '?')} "
                     f"T:{agent.stats_after.get('thirst', '?')} "
                     f"${agent.money_after}\n"
                 )
-            
-            if entry.deaths:
-                for death in entry.deaths:
-                    self._summary_file.write(f"  💀 {death['name']} died: {death['cause']}\n")
-            
-            self._summary_file.write("\n")
-            self._summary_file.flush()
+
+        if entry.deaths:
+            for death in entry.deaths:
+                summary_lines.append(f"  X {death['name']} died: {death['cause']}\n")
+
+        summary_lines.append("\n")
+        summary_data = "".join(summary_lines)
         
-        self.logger.debug(f"Turn {self._current_entry.turn} completed")
+        if self.async_logging and self._async_writer:
+            # Non-blocking write
+            self._async_writer.write(self._json_file_id, json_data)
+            self._async_writer.write(self._summary_file_id, summary_data)
+            self._items_queued += 2
+        else:
+            # Synchronous write
+            if self._json_file:
+                self._json_file.write(json_data)
+            if self._summary_file:
+                self._summary_file.write(summary_data)
+
+        # Batched flush - only flush every N turns
+        self._turns_since_flush += 1
+        if self._turns_since_flush >= self.flush_interval:
+            self._flush_files()
+            self._turns_since_flush = 0
+
+        # Clear current entry (don't accumulate in memory)
         self._current_entry = None
-    
+        self._agent_entry_map.clear()
+
+    def _flush_files(self) -> None:
+        """Flush file buffers to disk."""
+        if self.async_logging and self._async_writer:
+            self._async_writer.flush()
+        else:
+            if self._json_file:
+                self._json_file.flush()
+            if self._summary_file:
+                self._summary_file.flush()
+
     def close(self) -> None:
         """Close log files."""
-        if self._json_file:
-            self._json_file.close()
-            self._json_file = None
-        if self._summary_file:
-            self._summary_file.write(f"\nSession ended: {datetime.now()}\n")
-            self._summary_file.close()
-            self._summary_file = None
-        self.logger.info("Logging session closed")
-    
+        if self.async_logging and self._async_writer:
+            # Write final message before closing
+            if self._summary_file_id:
+                self._async_writer.write(
+                    self._summary_file_id,
+                    f"\nSession ended: {datetime.now()}\n"
+                )
+            
+            # Close files
+            if self._json_file_id:
+                self._async_writer.close_file(self._json_file_id)
+            if self._summary_file_id:
+                self._async_writer.close_file(self._summary_file_id)
+            
+            # Stop the writer thread
+            self._async_writer.stop()
+            self._async_writer = None
+        else:
+            if self._json_file:
+                self._json_file.close()
+                self._json_file = None
+            if self._summary_file:
+                self._summary_file.write(f"\nSession ended: {datetime.now()}\n")
+                self._summary_file.close()
+                self._summary_file = None
+
     def get_entries(self) -> list[TurnLogEntry]:
-        """Get all logged entries."""
-        return self._entries.copy()
+        """Get all logged entries.
+
+        Note: Returns empty list when logging optimized (entries not kept in memory).
+        """
+        return []
+    
+    def get_stats(self) -> dict:
+        """Get logging statistics."""
+        return {
+            "logging_enabled": self.logging_enabled,
+            "async_logging": self.async_logging,
+            "items_queued": self._items_queued,
+            "items_dropped": self._items_dropped,
+        }
 
 
 # Global logger instance
@@ -294,3 +565,11 @@ def reset_simulation_logger() -> SimulationLogger:
     _logger = SimulationLogger()
     return _logger
 
+
+# Ensure logger is closed on exit
+@atexit.register
+def _cleanup_logger():
+    global _logger
+    if _logger:
+        _logger.close()
+        _logger = None

backend/core/market.py

@@ -40,31 +40,32 @@ class Order:
     seller_id: str = ""
     resource_type: ResourceType = ResourceType.BERRIES
     quantity: int = 1
-    price_per_unit: int = 1
+    price_per_unit: int = 100  # Default to min_price from config
     created_turn: int = 0
     status: OrderStatus = OrderStatus.ACTIVE
-    
+
     # Price adjustment tracking
     turns_without_sale: int = 0
     original_price: int = 0
     last_adjusted_turn: int = 0  # Track when price was last changed
-    
+
     def __post_init__(self):
         if self.original_price == 0:
             self.original_price = self.price_per_unit
         if self.last_adjusted_turn == 0:
             self.last_adjusted_turn = self.created_turn
-    
+
     @property
     def total_price(self) -> int:
         """Get total price for all units."""
         return self.quantity * self.price_per_unit
-    
+
     def apply_discount(self, percentage: float = 0.1) -> None:
         """Apply a discount to the price."""
+        min_price = _get_min_price()
         reduction = max(1, int(self.price_per_unit * percentage))
-        self.price_per_unit = max(1, self.price_per_unit - reduction)
-    
+        self.price_per_unit = max(min_price, self.price_per_unit - reduction)
+
     def adjust_price(self, new_price: int, current_turn: int) -> bool:
         """Adjust the order's price. Returns True if successful."""
         if new_price < 1:
@@ -72,11 +73,11 @@ class Order:
         self.price_per_unit = new_price
         self.last_adjusted_turn = current_turn
         return True
-    
+
     def can_raise_price(self, current_turn: int, min_turns: int = 2) -> bool:
         """Check if enough time has passed to raise the price again."""
         return current_turn - self.last_adjusted_turn >= min_turns
-    
+
     def to_dict(self) -> dict:
         """Convert to dictionary for API serialization."""
         return {
@@ -104,7 +105,7 @@ class TradeResult:
     quantity: int = 0
     total_paid: int = 0
     message: str = ""
-    
+
     def to_dict(self) -> dict:
         return {
             "success": self.success,
@@ -124,31 +125,39 @@ def _get_market_config():
     return get_config().market
 
 
+def _get_min_price() -> int:
+    """Get minimum price floor from economy config."""
+    from backend.config import get_config
+    config = get_config()
+    economy = getattr(config, 'economy', None)
+    return getattr(economy, 'min_price', 100) if economy else 100
+
+
 @dataclass
 class OrderBook:
     """Central market order book with supply/demand tracking.
-    
+
     Features:
     - Track price history per resource type
     - Calculate supply/demand scores
     - Suggest prices based on market conditions
     - Allow sellers to adjust prices dynamically
-    
+
     Configuration is loaded from config.json.
     """
     orders: list[Order] = field(default_factory=list)
     trade_history: list[TradeResult] = field(default_factory=list)
     price_history: dict[ResourceType, PriceHistory] = field(default_factory=dict)
-    
+
     # Configuration - defaults loaded from config.json in __post_init__
     TURNS_BEFORE_DISCOUNT: int = 15
     DISCOUNT_RATE: float = 0.12
-    
+
     # Supply/demand thresholds
     LOW_SUPPLY_THRESHOLD: int = 3  # Less than this = scarcity
     HIGH_SUPPLY_THRESHOLD: int = 10  # More than this = surplus
     DEMAND_DECAY: float = 0.95  # How fast demand score decays per turn
-    
+
     def __post_init__(self):
         """Initialize price history and load config values."""
         # Load market config from config.json
@@ -158,11 +167,11 @@ class OrderBook:
             self.DISCOUNT_RATE = cfg.discount_rate
         except Exception:
             pass  # Use defaults if config not available
-        
+
         if not self.price_history:
             for resource_type in ResourceType:
                 self.price_history[resource_type] = PriceHistory()
-    
+
     def place_order(
         self,
         seller_id: str,
@@ -181,7 +190,7 @@ class OrderBook:
         )
         self.orders.append(order)
         return order
-    
+
     def cancel_order(self, order_id: str, seller_id: str) -> bool:
         """Cancel an order. Returns True if successful."""
         for order in self.orders:
@@ -190,11 +199,11 @@ class OrderBook:
                     order.status = OrderStatus.CANCELLED
                     return True
         return False
-    
+
     def get_active_orders(self) -> list[Order]:
         """Get all active orders."""
         return [o for o in self.orders if o.status == OrderStatus.ACTIVE]
-    
+
     def get_orders_by_type(self, resource_type: ResourceType) -> list[Order]:
         """Get all active orders for a specific resource type, sorted by price."""
         orders = [
@@ -202,19 +211,19 @@ class OrderBook:
             if o.status == OrderStatus.ACTIVE and o.resource_type == resource_type
         ]
         return sorted(orders, key=lambda o: o.price_per_unit)
-    
+
     def get_cheapest_order(self, resource_type: ResourceType) -> Optional[Order]:
         """Get the cheapest active order for a resource type."""
         orders = self.get_orders_by_type(resource_type)
         return orders[0] if orders else None
-    
+
     def get_orders_by_seller(self, seller_id: str) -> list[Order]:
         """Get all active orders from a specific seller."""
         return [
             o for o in self.orders
             if o.status == OrderStatus.ACTIVE and o.seller_id == seller_id
         ]
-    
+
     def cancel_seller_orders(self, seller_id: str) -> list[Order]:
         """Cancel all orders from a seller (e.g., when they die). Returns cancelled orders."""
         cancelled = []
@@ -223,7 +232,7 @@ class OrderBook:
                 order.status = OrderStatus.CANCELLED
                 cancelled.append(order)
         return cancelled
-    
+
     def execute_buy(
         self,
         buyer_id: str,
@@ -238,13 +247,13 @@ class OrderBook:
             if o.id == order_id and o.status == OrderStatus.ACTIVE:
                 order = o
                 break
-        
+
         if order is None:
             return TradeResult(
                 success=False,
                 message="Order not found or no longer active",
             )
-        
+
         # Check quantity
         actual_quantity = min(quantity, order.quantity)
         if actual_quantity <= 0:
@@ -252,7 +261,7 @@ class OrderBook:
                 success=False,
                 message="Invalid quantity",
             )
-        
+
         # Check buyer has enough money
         total_cost = actual_quantity * order.price_per_unit
         if buyer_money < total_cost:
@@ -265,12 +274,12 @@ class OrderBook:
                     message="Insufficient funds",
                 )
             total_cost = actual_quantity * order.price_per_unit
-        
+
         # Execute the trade
         order.quantity -= actual_quantity
         if order.quantity <= 0:
             order.status = OrderStatus.FILLED
-        
+
         result = TradeResult(
             success=True,
             order_id=order.id,
@@ -281,12 +290,15 @@ class OrderBook:
             total_paid=total_cost,
             message=f"Bought {actual_quantity} {order.resource_type.value} for {total_cost} coins",
         )
-        
+
         # Record sale for price history (we need current_turn but don't have it here)
         # The turn will be passed via the _record_sale call from engine
         self.trade_history.append(result)
+        # Keep trade history bounded to prevent memory growth
+        if len(self.trade_history) > 1000:
+            self.trade_history = self.trade_history[-500:]
         return result
-    
+
     def execute_multi_buy(
         self,
         buyer_id: str,
@@ -296,30 +308,30 @@ class OrderBook:
         """Execute multiple buy orders in one action. Returns list of trade results."""
         results = []
         remaining_money = buyer_money
-        
+
         for order_id, quantity in purchases:
             result = self.execute_buy(buyer_id, order_id, quantity, remaining_money)
             results.append(result)
             if result.success:
                 remaining_money -= result.total_paid
-        
+
         return results
-    
+
     def update_prices(self, current_turn: int) -> None:
         """Update order prices and supply/demand scores."""
         # Update supply/demand scores
         self._update_supply_demand_scores(current_turn)
-        
+
         # Apply automatic discounts to stale orders (keeping original behavior)
         for order in self.orders:
             if order.status != OrderStatus.ACTIVE:
                 continue
-            
+
             turns_waiting = current_turn - order.created_turn
             if turns_waiting > 0 and turns_waiting % self.TURNS_BEFORE_DISCOUNT == 0:
                 order.turns_without_sale = turns_waiting
                 order.apply_discount(self.DISCOUNT_RATE)
-    
+
     def _update_supply_demand_scores(self, current_turn: int) -> None:
         """Calculate current supply and demand scores for each resource."""
         for resource_type in ResourceType:
@@ -327,7 +339,7 @@ class OrderBook:
             if not history:
                 history = PriceHistory()
                 self.price_history[resource_type] = history
-            
+
             # Calculate supply score based on available quantity
             total_supply = self.get_total_supply(resource_type)
             if total_supply <= self.LOW_SUPPLY_THRESHOLD:
@@ -336,36 +348,36 @@ class OrderBook:
                 history.supply_score = min(1.0, 0.5 + (total_supply / self.HIGH_SUPPLY_THRESHOLD) * 0.5)
             else:
                 history.supply_score = 0.5
-            
+
             # Decay demand score over time
             history.demand_score *= self.DEMAND_DECAY
-    
+
     def get_total_supply(self, resource_type: ResourceType) -> int:
         """Get total quantity available for a resource type."""
         return sum(o.quantity for o in self.get_orders_by_type(resource_type))
-    
+
     def get_supply_demand_ratio(self, resource_type: ResourceType) -> float:
         """Get supply/demand ratio. <1 means scarcity, >1 means surplus."""
         history = self.price_history.get(resource_type, PriceHistory())
         demand = max(0.1, history.demand_score)
         supply = max(0.1, history.supply_score)
         return supply / demand
-    
+
     def get_suggested_price(self, resource_type: ResourceType, base_price: int) -> int:
         """Suggest a price based on supply/demand conditions.
-        
+
         Returns an adjusted price that accounts for market conditions:
         - Scarcity (low supply, high demand) -> higher price
         - Surplus (high supply, low demand) -> lower price
         """
         ratio = self.get_supply_demand_ratio(resource_type)
         history = self.price_history.get(resource_type, PriceHistory())
-        
+
         # Use average sale price as reference if available
         reference_price = base_price
         if history.avg_sale_price > 0:
             reference_price = int((base_price + history.avg_sale_price) / 2)
-        
+
         # Adjust based on supply/demand
         if ratio < 0.7:  # Scarcity - raise price
             price_multiplier = 1.0 + (0.7 - ratio) * 0.5  # Up to 35% increase
@@ -374,10 +386,10 @@ class OrderBook:
             price_multiplier = max(0.5, price_multiplier)  # Floor at 50%
         else:
             price_multiplier = 1.0
-        
+
         suggested = int(reference_price * price_multiplier)
-        return max(1, suggested)
-    
+        return max(_get_min_price(), suggested)
+
     def adjust_order_price(self, order_id: str, seller_id: str, new_price: int, current_turn: int) -> bool:
         """Adjust the price of an existing order. Returns True if successful."""
         for order in self.orders:
@@ -385,37 +397,37 @@ class OrderBook:
                 if order.status == OrderStatus.ACTIVE:
                     return order.adjust_price(new_price, current_turn)
         return False
-    
+
     def _record_sale(self, resource_type: ResourceType, price: int, quantity: int, current_turn: int) -> None:
         """Record a sale for price history tracking."""
         history = self.price_history.get(resource_type)
         if not history:
             history = PriceHistory()
             self.price_history[resource_type] = history
-        
+
         history.last_sale_price = price
         history.last_sale_turn = current_turn
-        
+
         # Update average sale price
         old_total = history.avg_sale_price * history.total_sold
         history.total_sold += quantity
         history.avg_sale_price = (old_total + price * quantity) / history.total_sold
-        
+
         # Increase demand score when sales happen
         history.demand_score = min(1.0, history.demand_score + 0.1 * quantity)
-    
+
     def cleanup_old_orders(self, max_age: int = 50) -> list[Order]:
         """Remove very old orders. Returns removed orders."""
         # For now, we don't auto-expire orders, but this could be enabled
         return []
-    
+
     def get_market_prices(self) -> dict[str, dict]:
         """Get current market price summary for each resource."""
         prices = {}
         for resource_type in ResourceType:
             orders = self.get_orders_by_type(resource_type)
             history = self.price_history.get(resource_type, PriceHistory())
-            
+
             if orders:
                 prices[resource_type.value] = {
                     "lowest_price": orders[0].price_per_unit,
@@ -437,7 +449,7 @@ class OrderBook:
                     "demand_score": round(history.demand_score, 2),
                 }
         return prices
-    
+
     def get_market_signal(self, resource_type: ResourceType) -> str:
         """Get a simple market signal for a resource: 'sell', 'hold', or 'buy'."""
         ratio = self.get_supply_demand_ratio(resource_type)
@@ -446,7 +458,7 @@ class OrderBook:
         elif ratio > 1.3:
             return "buy"   # Good time to buy - surplus
         return "hold"
-    
+
     def get_state_snapshot(self) -> dict:
         """Get market state for API."""
         return {

backend/core/storage.py

@@ -0,0 +1,450 @@
+"""State storage abstraction for simulation state.
+
+Provides a unified interface for storing and retrieving simulation state,
+with implementations for:
+- In-memory storage (default, fast but not persistent)
+- Redis storage (optional, enables decoupled UI polling and persistence)
+
+This allows the simulation loop to snapshot state without blocking,
+and enables external systems (like a web UI) to poll state independently.
+"""
+
+import json
+import time
+from abc import ABC, abstractmethod
+from dataclasses import dataclass
+from typing import Optional, Dict
+
+
+@dataclass
+class StateSnapshot:
+    """A point-in-time snapshot of simulation state."""
+    turn: int
+    timestamp: float
+    data: dict
+    
+    def to_json(self) -> str:
+        """Convert to JSON string."""
+        return json.dumps({
+            "turn": self.turn,
+            "timestamp": self.timestamp,
+            "data": self.data,
+        })
+    
+    @classmethod
+    def from_json(cls, json_str: str) -> "StateSnapshot":
+        """Create from JSON string."""
+        obj = json.loads(json_str)
+        return cls(
+            turn=obj["turn"],
+            timestamp=obj["timestamp"],
+            data=obj["data"],
+        )
+
+
+class StateStore(ABC):
+    """Abstract interface for state storage.
+    
+    Implementations should be thread-safe for concurrent read/write.
+    """
+    
+    @abstractmethod
+    def save_state(self, key: str, snapshot: StateSnapshot) -> bool:
+        """Save a state snapshot.
+        
+        Args:
+            key: Unique key for this state (e.g., "world", "market", "agent_123")
+            snapshot: The state snapshot to save
+            
+        Returns:
+            True if saved successfully
+        """
+        pass
+    
+    @abstractmethod
+    def get_state(self, key: str) -> Optional[StateSnapshot]:
+        """Get the latest state snapshot for a key.
+        
+        Args:
+            key: The state key to retrieve
+            
+        Returns:
+            The snapshot if found, None otherwise
+        """
+        pass
+    
+    @abstractmethod
+    def get_all_states(self, prefix: str = "") -> Dict[str, StateSnapshot]:
+        """Get all states matching a prefix.
+        
+        Args:
+            prefix: Key prefix to filter by (empty = all)
+            
+        Returns:
+            Dict mapping keys to snapshots
+        """
+        pass
+    
+    @abstractmethod
+    def delete_state(self, key: str) -> bool:
+        """Delete a state snapshot.
+        
+        Args:
+            key: The state key to delete
+            
+        Returns:
+            True if deleted (or didn't exist)
+        """
+        pass
+    
+    @abstractmethod
+    def clear_all(self) -> None:
+        """Clear all stored states."""
+        pass
+    
+    @abstractmethod
+    def is_healthy(self) -> bool:
+        """Check if the store is healthy and accessible."""
+        pass
+
+
+class MemoryStateStore(StateStore):
+    """In-memory state storage (default implementation).
+    
+    Fast but not persistent across restarts.
+    Thread-safe using a simple lock.
+    """
+    
+    def __init__(self, max_entries: int = 1000):
+        """Initialize memory store.
+        
+        Args:
+            max_entries: Maximum number of entries to keep (LRU eviction)
+        """
+        import threading
+        self._data: Dict[str, StateSnapshot] = {}
+        self._lock = threading.Lock()
+        self._max_entries = max_entries
+        self._access_order: list[str] = []  # For LRU tracking
+    
+    def save_state(self, key: str, snapshot: StateSnapshot) -> bool:
+        with self._lock:
+            # LRU eviction if at capacity
+            if len(self._data) >= self._max_entries and key not in self._data:
+                oldest = self._access_order.pop(0) if self._access_order else None
+                if oldest:
+                    self._data.pop(oldest, None)
+            
+            self._data[key] = snapshot
+            
+            # Update access order
+            if key in self._access_order:
+                self._access_order.remove(key)
+            self._access_order.append(key)
+            
+            return True
+    
+    def get_state(self, key: str) -> Optional[StateSnapshot]:
+        with self._lock:
+            snapshot = self._data.get(key)
+            if snapshot and key in self._access_order:
+                # Update access order for LRU
+                self._access_order.remove(key)
+                self._access_order.append(key)
+            return snapshot
+    
+    def get_all_states(self, prefix: str = "") -> Dict[str, StateSnapshot]:
+        with self._lock:
+            if not prefix:
+                return dict(self._data)
+            return {k: v for k, v in self._data.items() if k.startswith(prefix)}
+    
+    def delete_state(self, key: str) -> bool:
+        with self._lock:
+            self._data.pop(key, None)
+            if key in self._access_order:
+                self._access_order.remove(key)
+            return True
+    
+    def clear_all(self) -> None:
+        with self._lock:
+            self._data.clear()
+            self._access_order.clear()
+    
+    def is_healthy(self) -> bool:
+        return True
+
+
+class RedisStateStore(StateStore):
+    """Redis-backed state storage.
+    
+    Enables:
+    - Persistent state across restarts
+    - Decoupled UI polling (web clients can read state independently)
+    - Distributed access (multiple simulation instances)
+    
+    Requires redis-py: pip install redis
+    """
+    
+    def __init__(
+        self,
+        host: str = "localhost",
+        port: int = 6379,
+        db: int = 0,
+        password: Optional[str] = None,
+        prefix: str = "villsim:",
+        ttl_seconds: int = 3600,  # 1 hour default TTL
+    ):
+        """Initialize Redis store.
+        
+        Args:
+            host: Redis server host
+            port: Redis server port
+            db: Redis database number
+            password: Redis password (if required)
+            prefix: Key prefix for all keys (for namespacing)
+            ttl_seconds: Time-to-live for entries (0 = no expiry)
+        """
+        self._prefix = prefix
+        self._ttl = ttl_seconds
+        self._client = None
+        self._connection_params = {
+            "host": host,
+            "port": port,
+            "db": db,
+            "password": password,
+            "decode_responses": True,
+        }
+        self._connect()
+    
+    def _connect(self) -> None:
+        """Establish connection to Redis."""
+        try:
+            import redis
+            self._client = redis.Redis(**self._connection_params)
+            # Test connection
+            self._client.ping()
+        except ImportError:
+            raise ImportError(
+                "Redis support requires the 'redis' package. "
+                "Install with: pip install redis"
+            )
+        except Exception as e:
+            self._client = None
+            raise ConnectionError(f"Failed to connect to Redis: {e}")
+    
+    def _make_key(self, key: str) -> str:
+        """Create full Redis key with prefix."""
+        return f"{self._prefix}{key}"
+    
+    def save_state(self, key: str, snapshot: StateSnapshot) -> bool:
+        if not self._client:
+            return False
+        
+        try:
+            full_key = self._make_key(key)
+            data = snapshot.to_json()
+            
+            if self._ttl > 0:
+                self._client.setex(full_key, self._ttl, data)
+            else:
+                self._client.set(full_key, data)
+            
+            return True
+        except Exception:
+            return False
+    
+    def get_state(self, key: str) -> Optional[StateSnapshot]:
+        if not self._client:
+            return None
+        
+        try:
+            full_key = self._make_key(key)
+            data = self._client.get(full_key)
+            
+            if data:
+                return StateSnapshot.from_json(data)
+            return None
+        except Exception:
+            return None
+    
+    def get_all_states(self, prefix: str = "") -> Dict[str, StateSnapshot]:
+        if not self._client:
+            return {}
+        
+        try:
+            pattern = self._make_key(prefix + "*")
+            keys = self._client.keys(pattern)
+            
+            result = {}
+            for full_key in keys:
+                # Remove our prefix to get the original key
+                key = full_key[len(self._prefix):]
+                data = self._client.get(full_key)
+                if data:
+                    result[key] = StateSnapshot.from_json(data)
+            
+            return result
+        except Exception:
+            return {}
+    
+    def delete_state(self, key: str) -> bool:
+        if not self._client:
+            return False
+        
+        try:
+            full_key = self._make_key(key)
+            self._client.delete(full_key)
+            return True
+        except Exception:
+            return False
+    
+    def clear_all(self) -> None:
+        if not self._client:
+            return
+        
+        try:
+            pattern = self._make_key("*")
+            keys = self._client.keys(pattern)
+            if keys:
+                self._client.delete(*keys)
+        except Exception:
+            pass
+    
+    def is_healthy(self) -> bool:
+        if not self._client:
+            return False
+        
+        try:
+            self._client.ping()
+            return True
+        except Exception:
+            return False
+    
+    def publish_state_update(self, channel: str, key: str) -> None:
+        """Publish a state update notification (for real-time subscribers).
+        
+        This can be used for WebSocket-style updates where clients
+        subscribe to state changes.
+        """
+        if not self._client:
+            return
+        
+        try:
+            self._client.publish(
+                f"{self._prefix}updates:{channel}",
+                json.dumps({"key": key, "timestamp": time.time()})
+            )
+        except Exception:
+            pass
+
+
+class StubStateStore(StateStore):
+    """No-op state store for when storage is disabled.
+    
+    All operations succeed but don't actually store anything.
+    """
+    
+    def save_state(self, key: str, snapshot: StateSnapshot) -> bool:
+        return True
+    
+    def get_state(self, key: str) -> Optional[StateSnapshot]:
+        return None
+    
+    def get_all_states(self, prefix: str = "") -> Dict[str, StateSnapshot]:
+        return {}
+    
+    def delete_state(self, key: str) -> bool:
+        return True
+    
+    def clear_all(self) -> None:
+        pass
+    
+    def is_healthy(self) -> bool:
+        return True
+
+
+# Global state store instance
+_state_store: Optional[StateStore] = None
+
+
+def get_state_store() -> StateStore:
+    """Get the global state store instance.
+    
+    Creates a store based on config:
+    - If Redis is configured and available, uses Redis
+    - Otherwise falls back to in-memory storage
+    """
+    global _state_store
+    
+    if _state_store is None:
+        _state_store = _create_state_store()
+    
+    return _state_store
+
+
+def _create_state_store() -> StateStore:
+    """Create the appropriate state store based on config."""
+    from backend.config import get_config
+    
+    config = get_config()
+    
+    # Check for Redis config
+    redis_config = getattr(config, 'redis', None)
+    
+    if redis_config and getattr(redis_config, 'enabled', False):
+        try:
+            store = RedisStateStore(
+                host=getattr(redis_config, 'host', 'localhost'),
+                port=getattr(redis_config, 'port', 6379),
+                db=getattr(redis_config, 'db', 0),
+                password=getattr(redis_config, 'password', None),
+                prefix=getattr(redis_config, 'prefix', 'villsim:'),
+                ttl_seconds=getattr(redis_config, 'ttl_seconds', 3600),
+            )
+            if store.is_healthy():
+                return store
+        except Exception:
+            # Fall through to memory store
+            pass
+    
+    # Check if storage is disabled
+    perf_config = getattr(config, 'performance', None)
+    if perf_config and not getattr(perf_config, 'state_storage_enabled', True):
+        return StubStateStore()
+    
+    # Default to memory store
+    return MemoryStateStore()
+
+
+def reset_state_store() -> None:
+    """Reset the global state store."""
+    global _state_store
+    if _state_store:
+        _state_store.clear_all()
+    _state_store = None
+
+
+def save_simulation_state(turn: int, state_data: dict) -> bool:
+    """Convenience function to save simulation state.
+    
+    Args:
+        turn: Current simulation turn
+        state_data: Full state dict (world, market, agents, etc.)
+        
+    Returns:
+        True if saved successfully
+    """
+    store = get_state_store()
+    snapshot = StateSnapshot(
+        turn=turn,
+        timestamp=time.time(),
+        data=state_data,
+    )
+    return store.save_state("simulation:current", snapshot)
+
+
+def get_simulation_state() -> Optional[StateSnapshot]:
+    """Convenience function to get current simulation state."""
+    store = get_state_store()
+    return store.get_state("simulation:current")

backend/core/world.py

@@ -3,6 +3,8 @@
 The world spawns diverse agents with varied personality traits,
 skills, and starting conditions to create emergent professions
 and class inequality.
+
+Now includes age-based lifecycle with birth and death by old age.
 """
 
 import random
@@ -15,6 +17,7 @@ from backend.domain.personality import (
     PersonalityTraits, Skills,
     generate_random_personality, generate_random_skills
 )
+from backend.domain.resources import ResourceType
 
 
 class TimeOfDay(Enum):
@@ -32,7 +35,7 @@ def _get_world_config_from_file():
 @dataclass
 class WorldConfig:
     """Configuration for the world.
-    
+
     Default values are loaded from config.json via create_world_config().
     These hardcoded defaults are only fallbacks.
     """
@@ -64,11 +67,33 @@ class World:
     current_day: int = 1
     step_in_day: int = 0
     time_of_day: TimeOfDay = TimeOfDay.DAY
-    
+
+    # Agent index for O(1) lookups by ID
+    _agent_index: dict = field(default_factory=dict)
+
     # Statistics
     total_agents_spawned: int = 0
     total_agents_died: int = 0
-    
+    total_births: int = 0
+    total_deaths_by_age: int = 0
+    total_deaths_by_starvation: int = 0
+    total_deaths_by_thirst: int = 0
+    total_deaths_by_cold: int = 0
+
+    # Village-wide storage tracking (for storage limits)
+    village_storage: dict = field(default_factory=lambda: {
+        "meat": 0,
+        "berries": 0,
+        "water": 0,
+        "wood": 0,
+        "hide": 0,
+        "clothes": 0,
+    })
+
+    # Cached statistics (updated periodically for performance)
+    _cached_stats: Optional[dict] = field(default=None)
+    _stats_cache_turn: int = field(default=-1)
+
     def spawn_agent(
         self,
         name: Optional[str] = None,
@@ -76,30 +101,43 @@ class World:
         position: Optional[Position] = None,
         archetype: Optional[str] = None,
         starting_money: Optional[int] = None,
+        age: Optional[int] = None,
+        generation: int = 0,
+        parent_ids: Optional[list[str]] = None,
     ) -> Agent:
         """Spawn a new agent in the world with unique personality.
-        
+
         Args:
             name: Agent name (auto-generated if None)
             profession: Deprecated, now derived from personality
             position: Starting position (random if None)
             archetype: Personality archetype ("hunter", "gatherer", "trader", etc.)
             starting_money: Starting money (random with inequality if None)
+            age: Starting age (random within config range if None)
+            generation: 0 for initial spawn, 1+ for born in simulation
+            parent_ids: IDs of parent agents (for lineage tracking)
         """
         if position is None:
             position = Position(
                 x=random.randint(0, self.config.width - 1),
                 y=random.randint(0, self.config.height - 1),
             )
-        
-        # Generate unique personality and skills
+
+        # Get age config for age calculation
+        from backend.config import get_config
+        age_config = get_config().age
+
+        # Calculate starting age
+        if age is None:
+            age = random.randint(age_config.min_start_age, age_config.max_start_age)
+
+        # Generate unique personality and skills (skills influenced by age)
         personality = generate_random_personality(archetype)
-        skills = generate_random_skills(personality)
-        
+        skills = generate_random_skills(personality, age=age)
+
         # Variable starting money for class inequality
         # Some agents start with more, some with less
         if starting_money is None:
-            from backend.config import get_config
             base_money = get_config().world.starting_money
             # Random multiplier: 0.3x to 2.0x base money
             # This creates natural class inequality
@@ -108,7 +146,7 @@ class World:
             if personality.trade_preference > 1.3:
                 money_multiplier *= 1.5
             starting_money = int(base_money * money_multiplier)
-        
+
         agent = Agent(
             name=name or f"Villager_{self.total_agents_spawned + 1}",
             profession=Profession.VILLAGER,  # Will be updated based on personality
@@ -116,19 +154,190 @@ class World:
             personality=personality,
             skills=skills,
             money=starting_money,
+            age=age,
+            birth_day=self.current_day,
+            generation=generation,
+            parent_ids=parent_ids or [],
         )
-        
+
         self.agents.append(agent)
+        self._agent_index[agent.id] = agent  # Maintain index for O(1) lookups
         self.total_agents_spawned += 1
         return agent
-    
+
+    def spawn_child(self, parent: Agent) -> Optional[Agent]:
+        """Spawn a new agent as a child of an existing agent.
+
+        Birth chance is controlled by village prosperity (food abundance).
+        Parent transfers wealth to child at birth.
+
+        Returns the new agent or None if birth conditions not met.
+        """
+        from backend.config import get_config
+        age_config = get_config().age
+
+        # Check birth eligibility
+        if not parent.can_give_birth(self.current_day):
+            return None
+
+        # Calculate economy-based birth chance
+        # More food in village = higher birth rate
+        # But even in hard times, some births occur (base chance always applies)
+        prosperity = self.calculate_prosperity()
+
+        # Prosperity boosts birth rate: base_chance * (1 + prosperity * multiplier)
+        # At prosperity=0: birth_chance = base_chance
+        # At prosperity=1: birth_chance = base_chance * (1 + multiplier)
+        birth_chance = age_config.birth_base_chance * (1 + prosperity * age_config.birth_prosperity_multiplier)
+        birth_chance = min(0.20, birth_chance)  # Cap at 20%
+
+        if random.random() > birth_chance:
+            return None
+
+        # Birth happens! Child spawns near parent
+        child_pos = Position(
+            x=parent.position.x + random.uniform(-1, 1),
+            y=parent.position.y + random.uniform(-1, 1),
+        )
+        # Clamp to world bounds
+        child_pos.x = max(0, min(self.config.width - 1, child_pos.x))
+        child_pos.y = max(0, min(self.config.height - 1, child_pos.y))
+
+        # Child inherits some personality traits from parent with mutation
+        child_archetype = None  # Random, not determined by parent
+
+        # Wealth transfer: parent gives portion of their wealth to child
+        wealth_transfer = age_config.birth_wealth_transfer
+        child_money = int(parent.money * wealth_transfer)
+        parent.money -= child_money
+
+        # Ensure child has minimum viable wealth
+        min_child_money = int(get_config().world.starting_money * 0.3)
+        child_money = max(child_money, min_child_money)
+
+        # Child starts at configured age (adult)
+        child_age = age_config.child_start_age
+
+        child = self.spawn_agent(
+            name=f"Child_{self.total_agents_spawned + 1}",
+            position=child_pos,
+            archetype=child_archetype,
+            starting_money=child_money,
+            age=child_age,
+            generation=parent.generation + 1,
+            parent_ids=[parent.id],
+        )
+
+        # Parent also transfers some food to child
+        self._transfer_resources_to_child(parent, child)
+
+        # Record birth for parent
+        parent.record_birth(self.current_day, child.id)
+
+        self.total_births += 1
+        return child
+
+    def _transfer_resources_to_child(self, parent: Agent, child: Agent) -> None:
+        """Transfer some resources from parent to child at birth."""
+        # Transfer 1 of each food type parent has (if available)
+        for res_type in [ResourceType.MEAT, ResourceType.BERRIES, ResourceType.WATER]:
+            if parent.has_resource(res_type, 1):
+                parent.remove_from_inventory(res_type, 1)
+                from backend.domain.resources import Resource
+                child.add_to_inventory(Resource(
+                    type=res_type,
+                    quantity=1,
+                    created_turn=self.current_turn,
+                ))
+
+    def calculate_prosperity(self) -> float:
+        """Calculate village prosperity (0.0 to 1.0) based on food abundance.
+
+        Higher prosperity = more births allowed.
+        This creates population cycles tied to resource availability.
+        """
+        self.update_village_storage()
+        from backend.config import get_config
+        storage_config = get_config().storage
+
+        # Calculate how full each food storage is
+        meat_ratio = self.village_storage.get("meat", 0) / max(1, storage_config.village_meat_limit)
+        berries_ratio = self.village_storage.get("berries", 0) / max(1, storage_config.village_berries_limit)
+        water_ratio = self.village_storage.get("water", 0) / max(1, storage_config.village_water_limit)
+
+        # Average food abundance (weighted: meat most valuable)
+        prosperity = (meat_ratio * 0.4 + berries_ratio * 0.3 + water_ratio * 0.3)
+        return min(1.0, max(0.0, prosperity))
+
+    def process_inheritance(self, dead_agent: Agent) -> dict:
+        """Process inheritance when an agent dies.
+
+        Wealth and resources are distributed to living children.
+        If no children, wealth is distributed to random villagers (estate tax).
+
+        Returns dict with inheritance details.
+        """
+        from backend.config import get_config
+        age_config = get_config().age
+
+        if not age_config.inheritance_enabled:
+            return {"enabled": False}
+
+        inheritance_info = {
+            "enabled": True,
+            "deceased": dead_agent.name,
+            "total_money": dead_agent.money,
+            "total_resources": sum(r.quantity for r in dead_agent.inventory),
+            "beneficiaries": [],
+        }
+
+        # Find living children
+        living_children = []
+        for child_id in dead_agent.children_ids:
+            child = self.get_agent(child_id)
+            if child and child.is_alive():
+                living_children.append(child)
+
+        if living_children:
+            # Distribute equally among children
+            money_per_child = dead_agent.money // len(living_children)
+            for child in living_children:
+                child.money += money_per_child
+                inheritance_info["beneficiaries"].append({
+                    "name": child.name,
+                    "money": money_per_child,
+                })
+
+            # Distribute resources (round-robin)
+            for i, resource in enumerate(dead_agent.inventory):
+                recipient = living_children[i % len(living_children)]
+                recipient.add_to_inventory(resource)
+        else:
+            # No children - distribute to random villagers (estate tax effect)
+            living = self.get_living_agents()
+            if living:
+                # Give money to poorest villagers
+                poorest = sorted(living, key=lambda a: a.money)[:3]
+                if poorest:
+                    money_each = dead_agent.money // len(poorest)
+                    for villager in poorest:
+                        villager.money += money_each
+                        inheritance_info["beneficiaries"].append({
+                            "name": villager.name,
+                            "money": money_each,
+                            "relation": "community"
+                        })
+
+        # Clear dead agent's inventory (already distributed or lost)
+        dead_agent.inventory.clear()
+        dead_agent.money = 0
+
+        return inheritance_info
+
     def get_agent(self, agent_id: str) -> Optional[Agent]:
-        """Get an agent by ID."""
-        for agent in self.agents:
-            if agent.id == agent_id:
-                return agent
-        return None
-    
+        """Get an agent by ID (O(1) lookup via index)."""
+        return self._agent_index.get(agent_id)
+
     def remove_dead_agents(self) -> list[Agent]:
         """Remove all dead agents from the world. Returns list of removed agents.
         Note: This is now handled by the engine's corpse system for visualization.
@@ -136,44 +345,165 @@ class World:
         dead_agents = [a for a in self.agents if not a.is_alive() and not a.is_corpse()]
         # Don't actually remove here - let the engine handle corpse visualization
         return dead_agents
-    
-    def advance_time(self) -> None:
-        """Advance the simulation time by one step."""
+
+    def advance_time(self) -> bool:
+        """Advance the simulation time by one step.
+
+        Returns True if a new day started (for age/birth processing).
+        """
         self.current_turn += 1
         self.step_in_day += 1
-        
+
         total_steps = self.config.day_steps + self.config.night_steps
-        
+        new_day = False
+
         if self.step_in_day > total_steps:
             self.step_in_day = 1
             self.current_day += 1
-        
+            new_day = True
+
         # Determine time of day
         if self.step_in_day <= self.config.day_steps:
             self.time_of_day = TimeOfDay.DAY
         else:
             self.time_of_day = TimeOfDay.NIGHT
-    
+
+        return new_day
+
+    def process_new_day(self) -> dict:
+        """Process all new-day events: aging, births, sinks.
+
+        Returns a dict with events that happened.
+        """
+        events = {
+            "aged_agents": [],
+            "births": [],
+            "age_deaths": [],
+            "storage_decay": {},
+            "taxes_collected": 0,
+            "random_events": [],
+        }
+
+        # Age all living agents
+        for agent in self.get_living_agents():
+            agent.age_one_day()
+            events["aged_agents"].append(agent.id)
+
+        # Check for births (only from living agents after aging)
+        for agent in self.get_living_agents():
+            if agent.can_give_birth(self.current_day):
+                child = self.spawn_child(agent)
+                if child:
+                    events["births"].append({
+                        "parent_id": agent.id,
+                        "child_id": child.id,
+                        "child_name": child.name,
+                    })
+
+        return events
+
+    def update_village_storage(self) -> None:
+        """Update the village-wide storage tracking."""
+        # Reset counts
+        for key in self.village_storage:
+            self.village_storage[key] = 0
+
+        # Count all resources in agent inventories
+        for agent in self.get_living_agents():
+            for resource in agent.inventory:
+                res_type = resource.type.value
+                if res_type in self.village_storage:
+                    self.village_storage[res_type] += resource.quantity
+
+    def get_storage_limit(self, resource_type: str) -> int:
+        """Get the storage limit for a resource type."""
+        from backend.config import get_config
+        storage_config = get_config().storage
+
+        limit_map = {
+            "meat": storage_config.village_meat_limit,
+            "berries": storage_config.village_berries_limit,
+            "water": storage_config.village_water_limit,
+            "wood": storage_config.village_wood_limit,
+            "hide": storage_config.village_hide_limit,
+            "clothes": storage_config.village_clothes_limit,
+        }
+        return limit_map.get(resource_type, 999999)
+
+    def get_storage_available(self, resource_type: str) -> int:
+        """Get how much more of a resource can be stored village-wide."""
+        self.update_village_storage()
+        limit = self.get_storage_limit(resource_type)
+        current = self.village_storage.get(resource_type, 0)
+        return max(0, limit - current)
+
+    def is_storage_full(self, resource_type: str) -> bool:
+        """Check if village storage for a resource type is full."""
+        return self.get_storage_available(resource_type) <= 0
+
+    def record_death(self, agent: Agent, reason: str) -> None:
+        """Record a death and update statistics."""
+        self.total_agents_died += 1
+        if reason == "age":
+            self.total_deaths_by_age += 1
+        elif reason == "hunger":
+            self.total_deaths_by_starvation += 1
+        elif reason == "thirst":
+            self.total_deaths_by_thirst += 1
+        elif reason == "heat":
+            self.total_deaths_by_cold += 1
+
     def is_night(self) -> bool:
         """Check if it's currently night."""
         return self.time_of_day == TimeOfDay.NIGHT
-    
+
     def get_living_agents(self) -> list[Agent]:
         """Get all living agents (excludes corpses)."""
         return [a for a in self.agents if a.is_alive() and not a.is_corpse()]
-    
+
     def get_statistics(self) -> dict:
-        """Get current world statistics including wealth distribution."""
+        """Get current world statistics including wealth distribution and demographics.
+
+        Uses caching based on performance config to avoid recalculating every turn.
+        """
+        from backend.config import get_config
+        perf_config = get_config().performance
+
+        # Check if we can use cached stats
+        if (self._cached_stats is not None and
+            self.current_turn - self._stats_cache_turn < perf_config.stats_update_interval):
+            # Update just the essential changing values
+            self._cached_stats["current_turn"] = self.current_turn
+            self._cached_stats["current_day"] = self.current_day
+            self._cached_stats["step_in_day"] = self.step_in_day
+            self._cached_stats["time_of_day"] = self.time_of_day.value
+            return self._cached_stats
+
         living = self.get_living_agents()
         total_money = sum(a.money for a in living)
-        
+
         # Count emergent professions (updated based on current skills)
         profession_counts = {}
         for agent in living:
             agent._update_profession()  # Update based on current state
             prof = agent.profession.value
             profession_counts[prof] = profession_counts.get(prof, 0) + 1
-        
+
+        # Age demographics
+        age_distribution = {"young": 0, "prime": 0, "old": 0}
+        ages = []
+        generations = {}
+        for agent in living:
+            category = agent.get_age_category()
+            age_distribution[category] = age_distribution.get(category, 0) + 1
+            ages.append(agent.age)
+            gen = agent.generation
+            generations[gen] = generations.get(gen, 0) + 1
+
+        avg_age = sum(ages) / len(ages) if ages else 0
+        oldest_age = max(ages) if ages else 0
+        youngest_age = min(ages) if ages else 0
+
         # Calculate wealth inequality metrics
         if living:
             moneys = sorted([a.money for a in living])
@@ -181,17 +511,22 @@ class World:
             median_money = moneys[len(moneys) // 2]
             richest = moneys[-1] if moneys else 0
             poorest = moneys[0] if moneys else 0
-            
-            # Gini coefficient for inequality (0 = perfect equality, 1 = max inequality)
+
+            # Gini coefficient - O(n) algorithm instead of O(n²)
+            # Uses sorted list: Gini = (2 * sum(i * x_i)) / (n * sum(x_i)) - (n + 1) / n
             n = len(moneys)
             if n > 1 and total_money > 0:
-                sum_of_diffs = sum(abs(m1 - m2) for m1 in moneys for m2 in moneys)
-                gini = sum_of_diffs / (2 * n * total_money)
+                weighted_sum = sum((i + 1) * m for i, m in enumerate(moneys))
+                gini = (2 * weighted_sum) / (n * total_money) - (n + 1) / n
+                gini = max(0.0, min(1.0, gini))  # Clamp to [0, 1]
             else:
                 gini = 0
         else:
             avg_money = median_money = richest = poorest = gini = 0
-        
+
+        # Update village storage
+        self.update_village_storage()
+
         return {
             "current_turn": self.current_turn,
             "current_day": self.current_day,
@@ -200,6 +535,7 @@ class World:
             "living_agents": len(living),
             "total_agents_spawned": self.total_agents_spawned,
             "total_agents_died": self.total_agents_died,
+            "total_births": self.total_births,
             "total_money_in_circulation": total_money,
             "professions": profession_counts,
             # Wealth inequality metrics
@@ -208,8 +544,28 @@ class World:
             "richest_agent": richest,
             "poorest_agent": poorest,
             "gini_coefficient": round(gini, 3),
+            # Age demographics
+            "age_distribution": age_distribution,
+            "avg_age": round(avg_age, 1),
+            "oldest_agent": oldest_age,
+            "youngest_agent": youngest_age,
+            "generations": generations,
+            # Death statistics
+            "deaths_by_cause": {
+                "age": self.total_deaths_by_age,
+                "starvation": self.total_deaths_by_starvation,
+                "thirst": self.total_deaths_by_thirst,
+                "cold": self.total_deaths_by_cold,
+            },
+            # Village storage
+            "village_storage": self.village_storage.copy(),
         }
-    
+
+        # Cache the computed stats
+        self._cached_stats = stats
+        self._stats_cache_turn = self.current_turn
+        return stats
+
     def get_state_snapshot(self) -> dict:
         """Get a full snapshot of the world state for API."""
         return {
@@ -221,10 +577,10 @@ class World:
             "agents": [a.to_dict() for a in self.agents],
             "statistics": self.get_statistics(),
         }
-    
+
     def initialize(self) -> None:
         """Initialize the world with diverse starting agents.
-        
+
         Creates a mix of agent archetypes to seed profession diversity:
         - Some hunters (risk-takers who hunt)
         - Some gatherers (cautious resource collectors)
@@ -232,7 +588,7 @@ class World:
         - Some generalists (balanced approach)
         """
         n = self.config.initial_agents
-        
+
         # Distribute archetypes for diversity
         # ~15% hunters, ~15% gatherers, ~15% traders, ~10% woodcutters, 45% random
         archetypes = (
@@ -241,14 +597,14 @@ class World:
             ["trader"] * max(1, n // 7) +
             ["woodcutter"] * max(1, n // 10)
         )
-        
+
         # Fill remaining slots with random (no archetype)
         while len(archetypes) < n:
             archetypes.append(None)
-        
+
         # Shuffle to randomize positions
         random.shuffle(archetypes)
-        
+
         for archetype in archetypes:
             self.spawn_agent(archetype=archetype)
 

backend/domain/agent.py

@@ -25,6 +25,12 @@ def _get_agent_stats_config():
     return get_config().agent_stats
 
 
+def _get_age_config():
+    """Get age configuration from global config."""
+    from backend.config import get_config
+    return get_config().age
+
+
 class Profession(Enum):
     """Agent professions - now derived from personality and skills."""
     VILLAGER = "villager"
@@ -96,14 +102,24 @@ class AgentStats:
     # Critical threshold - loaded from config
     CRITICAL_THRESHOLD: float = field(default=0.25)
 
-    def apply_passive_decay(self, has_clothes: bool = False) -> None:
-        """Apply passive stat decay each turn."""
-        self.energy = max(0, self.energy - self.ENERGY_DECAY)
-        self.hunger = max(0, self.hunger - self.HUNGER_DECAY)
-        self.thirst = max(0, self.thirst - self.THIRST_DECAY)
+    def apply_passive_decay(self, has_clothes: bool = False, decay_modifier: float = 1.0) -> None:
+        """Apply passive stat decay each turn.
+
+        Args:
+            has_clothes: Whether agent has clothes (reduces heat decay)
+            decay_modifier: Age-based modifier (old agents decay faster)
+        """
+        energy_decay = int(self.ENERGY_DECAY * decay_modifier)
+        hunger_decay = int(self.HUNGER_DECAY * decay_modifier)
+        thirst_decay = int(self.THIRST_DECAY * decay_modifier)
+
+        self.energy = max(0, self.energy - energy_decay)
+        self.hunger = max(0, self.hunger - hunger_decay)
+        self.thirst = max(0, self.thirst - thirst_decay)
 
         # Clothes reduce heat loss by 50%
-        heat_decay = self.HEAT_DECAY // 2 if has_clothes else self.HEAT_DECAY
+        heat_decay = int(self.HEAT_DECAY * decay_modifier)
+        heat_decay = heat_decay // 2 if has_clothes else heat_decay
         self.heat = max(0, self.heat - heat_decay)
 
     def is_critical(self) -> bool:
@@ -217,6 +233,11 @@ class Agent:
     Stats, inventory slots, and starting money are loaded from config.json.
     Each agent now has unique personality traits and skills that create
     emergent behaviors and professions.
+
+    Age affects skills, energy costs, and survival:
+    - Young (< 25): Learning faster, lower skill effectiveness, less energy cost
+    - Prime (25-45): Peak performance
+    - Old (> 45): Higher skill effectiveness (wisdom), but higher energy costs
     """
     id: str = field(default_factory=lambda: str(uuid4())[:8])
     name: str = ""
@@ -230,6 +251,15 @@ class Agent:
     personality: PersonalityTraits = field(default_factory=PersonalityTraits)
     skills: Skills = field(default_factory=Skills)
 
+    # Age system - age is in "years" where 1 year = 1 simulation day
+    age: int = field(default=-1)  # -1 signals to use random start age
+    max_age: int = field(default=-1)  # -1 signals to calculate from config
+    birth_day: int = 0  # Day this agent was born (0 = initial spawn)
+    last_birth_day: int = -1000  # Last day this agent gave birth (for cooldown)
+    parent_ids: list[str] = field(default_factory=list)  # IDs of parents (for lineage)
+    children_ids: list[str] = field(default_factory=list)  # IDs of children
+    generation: int = 0  # 0 = initial spawn, 1+ = born in simulation
+
     # Movement and action tracking
     home_position: Position = field(default_factory=Position)
     current_action: AgentAction = field(default_factory=AgentAction)
@@ -267,6 +297,21 @@ class Agent:
         if self.INVENTORY_SLOTS == -1:
             self.INVENTORY_SLOTS = config.inventory_slots
 
+        # Initialize age system
+        age_config = _get_age_config()
+        if self.age == -1:
+            # Random starting age within configured range
+            self.age = random.randint(age_config.min_start_age, age_config.max_start_age)
+        if self.max_age == -1:
+            # Calculate max age with variance
+            variance = random.randint(-age_config.max_age_variance, age_config.max_age_variance)
+            self.max_age = age_config.base_max_age + variance
+
+        # Apply age-based max energy adjustment for old agents
+        if self.get_age_category() == "old":
+            self.stats.MAX_ENERGY = int(self.stats.MAX_ENERGY * age_config.old_max_energy_multiplier)
+            self.stats.energy = min(self.stats.energy, self.stats.MAX_ENERGY)
+
         # Update profession based on personality and skills
         self._update_profession()
 
@@ -282,6 +327,111 @@ class Agent:
         }
         self.profession = profession_map.get(prof_type, Profession.VILLAGER)
 
+    def get_age_category(self) -> str:
+        """Get the agent's age category: 'young', 'prime', or 'old'."""
+        age_config = _get_age_config()
+        if self.age < age_config.young_age_threshold:
+            return "young"
+        elif self.age <= age_config.old_age_threshold:
+            return "prime"
+        else:
+            return "old"
+
+    def get_skill_modifier(self) -> float:
+        """Get skill effectiveness modifier based on age.
+
+        Young agents are less effective but learn faster.
+        Old agents are more effective (wisdom) but learn slower.
+        """
+        age_config = _get_age_config()
+        category = self.get_age_category()
+        if category == "young":
+            return age_config.young_skill_multiplier
+        elif category == "prime":
+            return age_config.prime_skill_multiplier
+        else:
+            return age_config.old_skill_multiplier
+
+    def get_learning_modifier(self) -> float:
+        """Get learning rate modifier based on age."""
+        age_config = _get_age_config()
+        category = self.get_age_category()
+        if category == "young":
+            return age_config.young_learning_multiplier
+        elif category == "prime":
+            return age_config.prime_learning_multiplier
+        else:
+            return age_config.old_learning_multiplier
+
+    def get_energy_cost_modifier(self) -> float:
+        """Get energy cost modifier based on age.
+
+        Young agents use less energy.
+        Old agents use more energy.
+        """
+        age_config = _get_age_config()
+        category = self.get_age_category()
+        if category == "young":
+            return age_config.young_energy_cost_multiplier
+        elif category == "prime":
+            return age_config.prime_energy_cost_multiplier
+        else:
+            return age_config.old_energy_cost_multiplier
+
+    def get_decay_modifier(self) -> float:
+        """Get stat decay modifier based on age.
+
+        Old agents decay faster (frailer).
+        """
+        age_config = _get_age_config()
+        if self.get_age_category() == "old":
+            return age_config.old_decay_multiplier
+        return 1.0
+
+    def age_one_day(self) -> None:
+        """Age the agent by one day (called at day transition)."""
+        age_config = _get_age_config()
+        self.age += age_config.age_per_day
+
+        # Check if agent just became old - reduce max energy
+        if self.age == age_config.old_age_threshold + 1:
+            self.stats.MAX_ENERGY = int(self.stats.MAX_ENERGY * age_config.old_max_energy_multiplier)
+            self.stats.energy = min(self.stats.energy, self.stats.MAX_ENERGY)
+
+    def is_too_old(self) -> bool:
+        """Check if agent has exceeded their maximum age."""
+        return self.age >= self.max_age
+
+    def can_give_birth(self, current_day: int) -> bool:
+        """Check if agent is eligible to give birth."""
+        age_config = _get_age_config()
+
+        # Age check
+        if self.age < age_config.min_birth_age or self.age > age_config.max_birth_age:
+            return False
+
+        # Cooldown check
+        days_since_birth = current_day - self.last_birth_day
+        if days_since_birth < age_config.birth_cooldown_days:
+            return False
+
+        # Resource check
+        if self.stats.hunger < age_config.birth_food_requirement:
+            return False
+        if self.stats.energy < age_config.birth_energy_requirement:
+            return False
+
+        return True
+
+    def record_birth(self, current_day: int, child_id: str) -> None:
+        """Record that this agent gave birth."""
+        self.last_birth_day = current_day
+        self.children_ids.append(child_id)
+
+        # Birth is exhausting - reduce stats
+        self.stats.energy = max(0, self.stats.energy - 20)
+        self.stats.hunger = max(0, self.stats.hunger - 30)
+
     def record_action(self, action_type: str) -> None:
         """Record an action for profession tracking."""
         if action_type in self.actions_performed:
@@ -308,11 +458,13 @@ class Agent:
 
     def is_alive(self) -> bool:
         """Check if the agent is still alive."""
-        return (
-            self.stats.hunger > 0 and
-            self.stats.thirst > 0 and
-            self.stats.heat > 0
-        )
+        # Death by needs
+        if self.stats.hunger <= 0 or self.stats.thirst <= 0 or self.stats.heat <= 0:
+            return False
+        # Death by old age
+        if self.is_too_old():
+            return False
+        return True
 
     def is_corpse(self) -> bool:
         """Check if this agent is a corpse (died but still visible)."""
@@ -493,8 +645,9 @@ class Agent:
         return expired
 
     def apply_passive_decay(self) -> None:
-        """Apply passive stat decay for this turn."""
-        self.stats.apply_passive_decay(has_clothes=self.has_clothes())
+        """Apply passive stat decay for this turn, modified by age."""
+        decay_modifier = self.get_decay_modifier()
+        self.stats.apply_passive_decay(has_clothes=self.has_clothes(), decay_modifier=decay_modifier)
 
     def mark_dead(self, turn: int, reason: str) -> None:
         """Mark this agent as dead."""
@@ -522,6 +675,18 @@ class Agent:
             "last_action_result": self.last_action_result,
             "death_turn": self.death_turn,
             "death_reason": self.death_reason,
+            # Age system
+            "age": self.age,
+            "max_age": self.max_age,
+            "age_category": self.get_age_category(),
+            "birth_day": self.birth_day,
+            "generation": self.generation,
+            "parent_ids": self.parent_ids.copy(),
+            "children_count": len(self.children_ids),
+            # Age modifiers (for UI display)
+            "skill_modifier": round(self.get_skill_modifier(), 2),
+            "energy_cost_modifier": round(self.get_energy_cost_modifier(), 2),
+            "learning_modifier": round(self.get_learning_modifier(), 2),
             # New fields for agent diversity
             "personality": self.personality.to_dict(),
             "skills": self.skills.to_dict(),

backend/domain/personality.py

@@ -34,23 +34,23 @@ class ProfessionType(Enum):
 @dataclass
 class PersonalityTraits:
     """Unique personality traits that affect agent behavior.
-    
+
     These are set at birth and don't change during the agent's life.
     They create natural diversity in the population.
     """
     # How much the agent values accumulating wealth (0.1 = minimal, 0.9 = greedy)
     wealth_desire: float = 0.3
-    
+
     # How much the agent hoards resources vs trades them (0.1 = trades freely, 0.9 = hoards)
     hoarding_rate: float = 0.5
-    
+
     # Willingness to take risks (0.1 = very cautious, 0.9 = risk-taker)
     # Affects: hunting vs gathering preference, price decisions
     risk_tolerance: float = 0.5
-    
+
     # Sensitivity to good/bad deals (0.5 = not picky, 1.5 = very price conscious)
     price_sensitivity: float = 1.0
-    
+
     # Activity biases - how much the agent prefers each activity
     # Higher values = more likely to choose this activity
     # These create "profession tendencies"
@@ -58,12 +58,12 @@ class PersonalityTraits:
     gather_preference: float = 1.0    # Preference for gathering
     woodcut_preference: float = 1.0   # Preference for wood
     trade_preference: float = 1.0     # Preference for trading/market
-    
+
     # How social/market-oriented the agent is
     # High = frequent market visits, buys more from others
     # Low = self-sufficient, prefers to produce own resources
     market_affinity: float = 0.5
-    
+
     def to_dict(self) -> dict:
         return {
             "wealth_desire": round(self.wealth_desire, 2),
@@ -81,54 +81,63 @@ class PersonalityTraits:
 @dataclass
 class Skills:
     """Skills that improve with practice.
-    
+
     Each skill affects the outcome of related actions.
     Skills increase slowly through practice (use it or lose it).
     """
     # Combat/hunting skill - affects hunt success rate
     hunting: float = 1.0
-    
+
     # Foraging skill - affects gather output quantity
     gathering: float = 1.0
-    
+
     # Woodcutting skill - affects wood output
     woodcutting: float = 1.0
-    
+
     # Trading skill - affects prices (buy lower, sell higher)
     trading: float = 1.0
-    
+
     # Crafting skill - affects craft quality/success
     crafting: float = 1.0
-    
+
     # Skill improvement rate per action
     IMPROVEMENT_RATE: float = 0.02
-    
+
     # Skill decay rate per turn (use it or lose it, gentle decay)
     DECAY_RATE: float = 0.001
-    
+
     # Maximum skill level
     MAX_SKILL: float = 2.0
-    
+
     # Minimum skill level
     MIN_SKILL: float = 0.5
-    
-    def improve(self, skill_name: str, amount: Optional[float] = None) -> None:
-        """Improve a skill through practice."""
+
+    def improve(self, skill_name: str, amount: Optional[float] = None, learning_modifier: float = 1.0) -> None:
+        """Improve a skill through practice.
+
+        Args:
+            skill_name: Name of the skill to improve
+            amount: Base improvement amount (defaults to IMPROVEMENT_RATE)
+            learning_modifier: Age-based modifier (young learn faster, old learn slower)
+        """
         if amount is None:
             amount = self.IMPROVEMENT_RATE
-        
+
+        # Apply learning modifier (young agents learn faster)
+        amount = amount * learning_modifier
+
         if hasattr(self, skill_name):
             current = getattr(self, skill_name)
             new_value = min(self.MAX_SKILL, current + amount)
             setattr(self, skill_name, new_value)
-    
+
     def decay_all(self) -> None:
         """Apply gentle decay to all skills (use it or lose it)."""
         for skill_name in ['hunting', 'gathering', 'woodcutting', 'trading', 'crafting']:
             current = getattr(self, skill_name)
             new_value = max(self.MIN_SKILL, current - self.DECAY_RATE)
             setattr(self, skill_name, new_value)
-    
+
     def get_primary_skill(self) -> tuple[str, float]:
         """Get the agent's highest skill and its name."""
         skills = {
@@ -140,7 +149,7 @@ class Skills:
         }
         best_skill = max(skills, key=skills.get)
         return best_skill, skills[best_skill]
-    
+
     def to_dict(self) -> dict:
         return {
             "hunting": round(self.hunting, 3),
@@ -153,14 +162,14 @@ class Skills:
 
 def generate_random_personality(archetype: Optional[str] = None) -> PersonalityTraits:
     """Generate random personality traits.
-    
+
     If archetype is specified, traits will be biased towards that profession:
     - "hunter": High risk tolerance, high hunt preference
-    - "gatherer": Low risk tolerance, high gather preference  
+    - "gatherer": Low risk tolerance, high gather preference
     - "trader": High wealth desire, high market affinity, high trade preference
     - "hoarder": High hoarding rate, low market affinity
     - None: Fully random
-    
+
     Returns a PersonalityTraits instance with randomized values.
     """
     # Start with base random values
@@ -175,18 +184,18 @@ def generate_random_personality(archetype: Optional[str] = None) -> PersonalityT
         trade_preference=random.uniform(0.5, 1.5),
         market_affinity=random.uniform(0.2, 0.8),
     )
-    
+
     # Apply archetype biases
     if archetype == "hunter":
         traits.hunt_preference = random.uniform(1.3, 2.0)
         traits.risk_tolerance = random.uniform(0.6, 0.9)
         traits.gather_preference = random.uniform(0.3, 0.7)
-    
+
     elif archetype == "gatherer":
         traits.gather_preference = random.uniform(1.3, 2.0)
         traits.risk_tolerance = random.uniform(0.2, 0.5)
         traits.hunt_preference = random.uniform(0.3, 0.7)
-    
+
     elif archetype == "trader":
         traits.trade_preference = random.uniform(1.5, 2.5)
         traits.market_affinity = random.uniform(0.7, 0.95)
@@ -196,50 +205,63 @@ def generate_random_personality(archetype: Optional[str] = None) -> PersonalityT
         # Traders don't hunt/gather much
         traits.hunt_preference = random.uniform(0.2, 0.5)
         traits.gather_preference = random.uniform(0.2, 0.5)
-    
+
     elif archetype == "hoarder":
         traits.hoarding_rate = random.uniform(0.7, 0.95)
         traits.market_affinity = random.uniform(0.1, 0.4)
         traits.trade_preference = random.uniform(0.3, 0.7)
-    
+
     elif archetype == "woodcutter":
         traits.woodcut_preference = random.uniform(1.3, 2.0)
         traits.gather_preference = random.uniform(0.5, 0.8)
-    
+
     return traits
 
 
-def generate_random_skills(personality: PersonalityTraits) -> Skills:
-    """Generate starting skills influenced by personality.
-    
+def generate_random_skills(personality: PersonalityTraits, age: Optional[int] = None) -> Skills:
+    """Generate starting skills influenced by personality and age.
+
     Agents with strong preferences start with slightly better skills
     in those areas (natural talent).
+
+    Older agents start with higher skills (life experience).
     """
     # Base skill level with small random variation
     base = 1.0
     variance = 0.15
-    
+
+    # Age bonus: older agents have more experience
+    age_bonus = 0.0
+    if age is not None:
+        # Young agents (< 25): no bonus
+        # Prime agents (25-45): small bonus
+        # Old agents (> 45): larger bonus (wisdom)
+        if age >= 45:
+            age_bonus = 0.3 + random.uniform(0, 0.2)
+        elif age >= 25:
+            age_bonus = (age - 25) * 0.01 + random.uniform(0, 0.1)
+
     skills = Skills(
-        hunting=base + random.uniform(-variance, variance) + (personality.hunt_preference - 1.0) * 0.1,
-        gathering=base + random.uniform(-variance, variance) + (personality.gather_preference - 1.0) * 0.1,
-        woodcutting=base + random.uniform(-variance, variance) + (personality.woodcut_preference - 1.0) * 0.1,
-        trading=base + random.uniform(-variance, variance) + (personality.trade_preference - 1.0) * 0.1,
-        crafting=base + random.uniform(-variance, variance),
+        hunting=base + random.uniform(-variance, variance) + (personality.hunt_preference - 1.0) * 0.1 + age_bonus,
+        gathering=base + random.uniform(-variance, variance) + (personality.gather_preference - 1.0) * 0.1 + age_bonus,
+        woodcutting=base + random.uniform(-variance, variance) + (personality.woodcut_preference - 1.0) * 0.1 + age_bonus,
+        trading=base + random.uniform(-variance, variance) + (personality.trade_preference - 1.0) * 0.1 + age_bonus,
+        crafting=base + random.uniform(-variance, variance) + age_bonus,
     )
-    
+
     # Clamp all skills to valid range
     skills.hunting = max(skills.MIN_SKILL, min(skills.MAX_SKILL, skills.hunting))
     skills.gathering = max(skills.MIN_SKILL, min(skills.MAX_SKILL, skills.gathering))
     skills.woodcutting = max(skills.MIN_SKILL, min(skills.MAX_SKILL, skills.woodcutting))
     skills.trading = max(skills.MIN_SKILL, min(skills.MAX_SKILL, skills.trading))
     skills.crafting = max(skills.MIN_SKILL, min(skills.MAX_SKILL, skills.crafting))
-    
+
     return skills
 
 
 def determine_profession(personality: PersonalityTraits, skills: Skills) -> ProfessionType:
     """Determine an agent's emergent profession based on traits and skills.
-    
+
     This is for display/statistics - it doesn't affect behavior directly.
     The behavior is determined by the traits and skills themselves.
     """
@@ -250,27 +272,27 @@ def determine_profession(personality: PersonalityTraits, skills: Skills) -> Prof
         ProfessionType.WOODCUTTER: personality.woodcut_preference * skills.woodcutting,
         ProfessionType.TRADER: personality.trade_preference * skills.trading * personality.market_affinity * 1.5,
     }
-    
+
     # Find the best match
     best_profession = max(scores, key=scores.get)
     best_score = scores[best_profession]
-    
+
     # If no clear winner (all scores similar), they're a generalist
     second_best = sorted(scores.values(), reverse=True)[1]
     if best_score < second_best * 1.2:
         return ProfessionType.GENERALIST
-    
+
     return best_profession
 
 
 def get_action_skill_modifier(skill_value: float) -> float:
     """Convert skill value to action modifier.
-    
+
     Skill 0.5 = 0.75x effectiveness
-    Skill 1.0 = 1.0x effectiveness  
+    Skill 1.0 = 1.0x effectiveness
     Skill 1.5 = 1.25x effectiveness
     Skill 2.0 = 1.5x effectiveness
-    
+
     This creates meaningful but not overpowering differences.
     """
     # Linear scaling: (skill - 1.0) * 0.5 + 1.0
@@ -280,16 +302,16 @@ def get_action_skill_modifier(skill_value: float) -> float:
 
 def get_trade_price_modifier(skill_value: float, is_buying: bool) -> float:
     """Get price modifier for trading based on skill.
-    
+
     Higher trading skill = better deals:
     - When buying: lower prices (modifier < 1)
     - When selling: higher prices (modifier > 1)
-    
+
     Skill 1.0 = no modifier
     Skill 2.0 = 15% better deals
     """
     modifier = (skill_value - 1.0) * 0.15
-    
+
     if is_buying:
         return max(0.85, 1.0 - modifier)  # Lower is better for buying
     else:

backend/main.py

@@ -4,6 +4,7 @@ import os
 import uvicorn
 from fastapi import FastAPI
 from fastapi.middleware.cors import CORSMiddleware
+from fastapi.responses import RedirectResponse
 from fastapi.staticfiles import StaticFiles
 
 from backend.api.routes import router
@@ -48,14 +49,8 @@ async def startup_event():
 
 @app.get("/", tags=["root"])
 def root():
-    """Root endpoint with API information."""
-    return {
-        "name": "Village Simulation API",
-        "version": "1.0.0",
-        "docs": "/docs",
-        "web_frontend": "/web/",
-        "status": "running",
-    }
+    """Root endpoint - redirect to web frontend."""
+    return RedirectResponse(url="/web/")
 
 
 @app.get("/health", tags=["health"])
@@ -69,12 +64,19 @@ def health_check():
     }
 
 
-# ============== Web Frontend Static Files ==============
+# ============== Web Frontend ==============
 
-# Mount static files for web frontend
-# Access at http://localhost:8000/web/
+@app.get("/web", include_in_schema=False)
+def redirect_to_web_frontend():
+    """Redirect /web to /web/ for static file serving."""
+    return RedirectResponse(url="/web/")
+
+
+# Mount static files for web frontend (access at http://localhost:8000/web/)
 if os.path.exists(WEB_FRONTEND_PATH):
     app.mount("/web", StaticFiles(directory=WEB_FRONTEND_PATH, html=True), name="web_frontend")
+else:
+    print(f"Warning: Web frontend not found at {WEB_FRONTEND_PATH}")
 
 
 def main():

config.json

@@ -1,51 +1,133 @@
 {
+  "performance": {
+    "logging_enabled": false,
+    "detailed_logging": false,
+    "async_logging": true,
+    "log_flush_interval": 50,
+    "max_turn_logs": 100,
+    "stats_update_interval": 10,
+    "state_storage_enabled": true
+  },
+  "ai": {
+    "goap_max_iterations": 30,
+    "goap_max_plan_depth": 2,
+    "reactive_fallback": true,
+    "use_bdi": true
+  },
+  "bdi": {
+    "thinking_interval": 1,
+    "max_consecutive_failures": 2,
+    "priority_switch_threshold": 1.5,
+    "memory_max_events": 50,
+    "memory_decay_rate": 0.1
+  },
+  "redis": {
+    "enabled": false,
+    "host": "localhost",
+    "port": 6379,
+    "db": 0,
+    "password": null,
+    "prefix": "villsim:",
+    "ttl_seconds": 3600
+  },
   "agent_stats": {
     "max_energy": 50,
     "max_hunger": 100,
     "max_thirst": 100,
     "max_heat": 100,
     "start_energy": 50,
-    "start_hunger": 70,
-    "start_thirst": 75,
+    "start_hunger": 80,
+    "start_thirst": 85,
     "start_heat": 100,
     "energy_decay": 1,
     "hunger_decay": 2,
-    "thirst_decay": 3,
-    "heat_decay": 3,
+    "thirst_decay": 2,
+    "heat_decay": 2,
     "critical_threshold": 0.25,
     "low_energy_threshold": 12
   },
+  "age": {
+    "min_start_age": 18,
+    "max_start_age": 28,
+    "young_age_threshold": 25,
+    "prime_age_start": 25,
+    "prime_age_end": 50,
+    "old_age_threshold": 50,
+    "base_max_age": 75,
+    "max_age_variance": 8,
+    "age_per_day": 1,
+    "birth_cooldown_days": 8,
+    "min_birth_age": 20,
+    "max_birth_age": 50,
+    "birth_base_chance": 0.06,
+    "birth_prosperity_multiplier": 2.5,
+    "birth_food_requirement": 40,
+    "birth_energy_requirement": 15,
+    "birth_wealth_transfer": 0.15,
+    "inheritance_enabled": true,
+    "child_start_age": 18,
+    "young_skill_multiplier": 0.85,
+    "young_learning_multiplier": 1.3,
+    "young_energy_cost_multiplier": 0.9,
+    "prime_skill_multiplier": 1.0,
+    "prime_learning_multiplier": 1.0,
+    "prime_energy_cost_multiplier": 1.0,
+    "old_skill_multiplier": 1.1,
+    "old_learning_multiplier": 0.7,
+    "old_energy_cost_multiplier": 1.15,
+    "old_max_energy_multiplier": 0.8,
+    "old_decay_multiplier": 1.1
+  },
+  "storage": {
+    "village_meat_limit": 200,
+    "village_berries_limit": 300,
+    "village_water_limit": 400,
+    "village_wood_limit": 400,
+    "village_hide_limit": 150,
+    "village_clothes_limit": 100,
+    "market_order_limit_per_agent": 5,
+    "market_total_order_limit": 500
+  },
+  "sinks": {
+    "daily_village_decay_rate": 0.01,
+    "daily_tax_rate": 0.005,
+    "random_event_chance": 0.02,
+    "fire_event_resource_loss": 0.05,
+    "theft_event_money_loss": 0.03,
+    "clothes_maintenance_per_day": 1,
+    "fire_wood_cost_per_night": 1
+  },
   "resources": {
-    "meat_decay": 10,
-    "berries_decay": 6,
-    "clothes_decay": 20,
-    "meat_hunger": 35,
-    "meat_energy": 12,
+    "meat_decay": 12,
+    "berries_decay": 8,
+    "clothes_decay": 30,
+    "meat_hunger": 45,
+    "meat_energy": 15,
     "berries_hunger": 10,
-    "berries_thirst": 4,
+    "berries_thirst": 3,
     "water_thirst": 50,
-    "fire_heat": 20
+    "fire_heat": 25
   },
   "actions": {
     "sleep_energy": 55,
     "rest_energy": 12,
-    "hunt_energy": -7,
+    "hunt_energy": -5,
     "gather_energy": -3,
-    "chop_wood_energy": -6,
+    "chop_wood_energy": -5,
     "get_water_energy": -2,
-    "weave_energy": -6,
-    "build_fire_energy": -4,
+    "weave_energy": -5,
+    "build_fire_energy": -3,
     "trade_energy": -1,
-    "hunt_success": 0.70,
-    "chop_wood_success": 0.90,
+    "hunt_success": 0.85,
+    "chop_wood_success": 0.9,
     "hunt_meat_min": 2,
     "hunt_meat_max": 5,
     "hunt_hide_min": 0,
     "hunt_hide_max": 2,
-    "gather_min": 2,
-    "gather_max": 4,
-    "chop_wood_min": 1,
-    "chop_wood_max": 3
+    "gather_min": 3,
+    "gather_max": 5,
+    "chop_wood_min": 2,
+    "chop_wood_max": 4
   },
   "world": {
     "width": 25,
@@ -53,8 +135,8 @@
     "initial_agents": 25,
     "day_steps": 10,
     "night_steps": 1,
-    "inventory_slots": 12,
-    "starting_money": 80
+    "inventory_slots": 15,
+    "starting_money": 8000
   },
   "market": {
     "turns_before_discount": 15,
@@ -62,10 +144,11 @@
     "base_price_multiplier": 1.3
   },
   "economy": {
-    "energy_to_money_ratio": 1.5,
+    "energy_to_money_ratio": 150,
+    "min_price": 100,
     "wealth_desire": 0.35,
     "buy_efficiency_threshold": 0.75,
-    "min_wealth_target": 50,
+    "min_wealth_target": 5000,
     "max_price_markup": 2.5,
     "min_price_discount": 0.4
   },

config_goap_optimized.json

@@ -0,0 +1,130 @@
+{
+  "ai": {
+    "goap_max_iterations": 50,
+    "goap_max_plan_depth": 3,
+    "reactive_fallback": true
+  },
+  "agent_stats": {
+    "max_energy": 50,
+    "max_hunger": 100,
+    "max_thirst": 100,
+    "max_heat": 100,
+    "start_energy": 50,
+    "start_hunger": 80,
+    "start_thirst": 85,
+    "start_heat": 100,
+    "energy_decay": 1,
+    "hunger_decay": 2,
+    "thirst_decay": 2,
+    "heat_decay": 2,
+    "critical_threshold": 0.25,
+    "low_energy_threshold": 12
+  },
+  "age": {
+    "min_start_age": 18,
+    "max_start_age": 28,
+    "young_age_threshold": 25,
+    "prime_age_start": 25,
+    "prime_age_end": 50,
+    "old_age_threshold": 50,
+    "base_max_age": 75,
+    "max_age_variance": 8,
+    "age_per_day": 1,
+    "birth_cooldown_days": 8,
+    "min_birth_age": 20,
+    "max_birth_age": 50,
+    "birth_base_chance": 0.06,
+    "birth_prosperity_multiplier": 2.5,
+    "birth_food_requirement": 40,
+    "birth_energy_requirement": 15,
+    "birth_wealth_transfer": 0.15,
+    "inheritance_enabled": true,
+    "child_start_age": 18,
+    "young_skill_multiplier": 0.85,
+    "young_learning_multiplier": 1.3,
+    "young_energy_cost_multiplier": 0.9,
+    "prime_skill_multiplier": 1.0,
+    "prime_learning_multiplier": 1.0,
+    "prime_energy_cost_multiplier": 1.0,
+    "old_skill_multiplier": 1.1,
+    "old_learning_multiplier": 0.7,
+    "old_energy_cost_multiplier": 1.15,
+    "old_max_energy_multiplier": 0.8,
+    "old_decay_multiplier": 1.1
+  },
+  "storage": {
+    "village_meat_limit": 200,
+    "village_berries_limit": 300,
+    "village_water_limit": 400,
+    "village_wood_limit": 400,
+    "village_hide_limit": 150,
+    "village_clothes_limit": 100,
+    "market_order_limit_per_agent": 5,
+    "market_total_order_limit": 500
+  },
+  "sinks": {
+    "daily_village_decay_rate": 0.01,
+    "daily_tax_rate": 0.005,
+    "random_event_chance": 0.02,
+    "fire_event_resource_loss": 0.05,
+    "theft_event_money_loss": 0.03,
+    "clothes_maintenance_per_day": 1,
+    "fire_wood_cost_per_night": 1
+  },
+  "resources": {
+    "meat_decay": 12,
+    "berries_decay": 8,
+    "clothes_decay": 30,
+    "meat_hunger": 45,
+    "meat_energy": 15,
+    "berries_hunger": 10,
+    "berries_thirst": 3,
+    "water_thirst": 50,
+    "fire_heat": 25
+  },
+  "actions": {
+    "sleep_energy": 55,
+    "rest_energy": 12,
+    "hunt_energy": -5,
+    "gather_energy": -3,
+    "chop_wood_energy": -5,
+    "get_water_energy": -2,
+    "weave_energy": -5,
+    "build_fire_energy": -3,
+    "trade_energy": -1,
+    "hunt_success": 0.85,
+    "chop_wood_success": 0.9,
+    "hunt_meat_min": 2,
+    "hunt_meat_max": 5,
+    "hunt_hide_min": 0,
+    "hunt_hide_max": 2,
+    "gather_min": 3,
+    "gather_max": 5,
+    "chop_wood_min": 2,
+    "chop_wood_max": 4
+  },
+  "world": {
+    "width": 25,
+    "height": 25,
+    "initial_agents": 25,
+    "day_steps": 10,
+    "night_steps": 1,
+    "inventory_slots": 15,
+    "starting_money": 80
+  },
+  "market": {
+    "turns_before_discount": 15,
+    "discount_rate": 0.12,
+    "base_price_multiplier": 1.3
+  },
+  "economy": {
+    "energy_to_money_ratio": 1.5,
+    "min_price": 1,
+    "wealth_desire": 0.35,
+    "buy_efficiency_threshold": 0.75,
+    "min_wealth_target": 50,
+    "max_price_markup": 2.5,
+    "min_price_discount": 0.4
+  },
+  "auto_step_interval": 0.15
+}

docs/design/simple-architecture.md

@@ -5,10 +5,10 @@ This document outlines the architecture for the Village Simulation based on [Vil
 ## 1. System Overview
 
 The system consists of two distinct applications communicating via HTTP (REST API):
-1.  **Backend (Server)**: Responsible for the entire simulation state, economic logic, AI decision-making, and turn management.
-2.  **Frontend (Client)**: A "dumb" terminal using **Pygame** that queries the current state to render it and sends user commands (if any) to the server.
+1.  **Backend (Server)**: Responsible for the entire simulation state, economic logic, AI decision-making (GOAP-based), and turn management.
+2.  **Frontend (Client)**: A web-based frontend (HTML/JavaScript) that queries the current state to render it and sends user commands to the server.
 
-This separation allows replacing the Pygame frontend with Web (React/Vue) or Unity in the future without changing the backend logic.
+This separation allows replacing the web frontend with other technologies (React/Vue, Unity, etc.) without changing the backend logic.
 
 ---
 
@@ -54,21 +54,24 @@ backend/
 
 ---
 
-## 3. Frontend Architecture (Pygame)
+## 3. Frontend Architecture (Web)
 
 The frontend acts as a **Visualizer**. It does not calculate simulation logic.
 
 ### 3.1. Structure
 
 ```text
-frontend/
-├── main.py                 # Pygame Game Loop
-├── client.py               # Network Client (requests lib)
-├── assets/                 # Sprites/Fonts
-└── renderer/               # Drawing Logic
-    ├── map_renderer.py     # Draws the grid/terrain
-    ├── agent_renderer.py   # Draws agents and their status bars
-    └── ui_renderer.py      # Draws text info (Market prices, Day/Night)
+web_frontend/
+├── index.html              # Main HTML page
+├── goap_debug.html         # GOAP debugging view
+├── styles.css              # Styling
+└── src/
+    ├── main.js             # Application entry point
+    ├── api.js              # Network client (fetch API)
+    ├── constants.js        # Configuration constants
+    └── scenes/             # Game scenes (Phaser.js)
+        ├── BootScene.js    # Loading scene
+        └── GameScene.js    # Main game visualization
 ```
 
 ### 3.2. Flow
@@ -77,12 +80,11 @@ frontend/
     *   Call `GET http://localhost:8000/state`.
     *   Receive JSON: `{"turn": 5, "time_of_day": "day", "agents": [...], "market": [...]}`.
 2.  **Update Step**:
-    *   Parse JSON into local simplified objects.
+    *   Parse JSON into JavaScript objects.
 3.  **Draw Step**:
-    *   Clear screen.
+    *   Update Phaser.js game scene.
     *   Render Agents at their coordinates.
     *   Render UI overlays (e.g., "Day 1, Step 5", "Total Coins: 500").
-    *   `pygame.display.flip()`.
 
 ---
 
@@ -97,12 +99,10 @@ Since the simulation involves AI agents acting autonomously, the Frontend is pri
     *   Frontend updates the screen.
 
 ### 4.1. The "God Mode" Problem
-To test the simulation efficiently, the Server will expose a **Simulation Controller**:
-*   **Manual Mode**: The server waits for a `POST /next_step` call to advance. The User presses `SPACE` in Pygame -> Pygame sends request -> Server updates -> Pygame fetches new state.
+To test the simulation efficiently, the Server exposes a **Simulation Controller**:
+*   **Manual Mode**: The server waits for a `POST /next_step` call to advance. The User clicks the advance button in the web frontend -> Frontend sends request -> Server updates -> Frontend fetches new state.
 *   **Auto Mode**: Server runs a background thread updating every N seconds. Frontend just polls.
 
-*Recommended for MVP: Manual Mode (Spacebar to advance turn).*
-
 ---
 
 ## 5. Technology Stack
@@ -110,12 +110,13 @@ To test the simulation efficiently, the Server will expose a **Simulation Contro
 *   **Language**: Python 3.11+
 *   **Backend Framework**: FastAPI (for speed and auto-generated docs).
 *   **Data Validation**: Pydantic.
-*   **Frontend**: Pygame Community Edition (pygame-ce).
-*   **Communication**: HTTP (Requests/Uvicorn).
+*   **AI System**: GOAP (Goal-Oriented Action Planning).
+*   **Frontend**: HTML/JavaScript with Phaser.js for rendering.
+*   **Communication**: HTTP (Fetch API/Uvicorn).
 
 ## 6. Future Extensibility (Why this architecture?)
 
-*   **Switch to Web**: Replace `frontend/` folder with a React app. The React app simply calls the same `GET /state` endpoint.
+*   **Switch to React/Vue**: Replace `web_frontend/` folder with a React app. The React app simply calls the same `GET /state` endpoint.
 *   **Switch to Unity**: Unity `UnityWebRequest` calls `GET /state`.
 *   **Database**: Currently state is in-memory (`core/engine.py`). Easy to swap for SQLite/Postgres later by adding a `repository` layer in Backend.
 

frontend/__init__.py

@@ -1,2 +0,0 @@
-"""Frontend package for Village Simulation visualization."""
-

frontend/client.py

@@ -1,180 +0,0 @@
-"""HTTP client for communicating with the Village Simulation backend."""
-
-import time
-from dataclasses import dataclass
-from typing import Optional, Any
-
-import requests
-from requests.exceptions import RequestException
-
-
-@dataclass
-class SimulationState:
-    """Parsed simulation state from the API."""
-    turn: int
-    day: int
-    step_in_day: int
-    time_of_day: str
-    world_width: int
-    world_height: int
-    agents: list[dict]
-    market_orders: list[dict]
-    market_prices: dict
-    statistics: dict
-    mode: str
-    is_running: bool
-    recent_logs: list[dict]
-    
-    @classmethod
-    def from_api_response(cls, data: dict) -> "SimulationState":
-        """Create from API response data."""
-        return cls(
-            turn=data.get("turn", 0),
-            day=data.get("day", 1),
-            step_in_day=data.get("step_in_day", 0),
-            time_of_day=data.get("time_of_day", "day"),
-            world_width=data.get("world_size", {}).get("width", 20),
-            world_height=data.get("world_size", {}).get("height", 20),
-            agents=data.get("agents", []),
-            market_orders=data.get("market", {}).get("orders", []),
-            market_prices=data.get("market", {}).get("prices", {}),
-            statistics=data.get("statistics", {}),
-            mode=data.get("mode", "manual"),
-            is_running=data.get("is_running", False),
-            recent_logs=data.get("recent_logs", []),
-        )
-    
-    def get_living_agents(self) -> list[dict]:
-        """Get only living agents."""
-        return [a for a in self.agents if a.get("is_alive", False)]
-
-
-class SimulationClient:
-    """HTTP client for the Village Simulation backend."""
-    
-    def __init__(self, base_url: str = "http://localhost:8000"):
-        self.base_url = base_url.rstrip("/")
-        self.api_url = f"{self.base_url}/api"
-        self.session = requests.Session()
-        self.last_state: Optional[SimulationState] = None
-        self.connected = False
-        self._retry_count = 0
-        self._max_retries = 3
-    
-    def _request(
-        self,
-        method: str,
-        endpoint: str,
-        json: Optional[dict] = None,
-        timeout: float = 5.0,
-    ) -> Optional[dict]:
-        """Make an HTTP request to the API."""
-        url = f"{self.api_url}{endpoint}"
-        
-        try:
-            response = self.session.request(
-                method=method,
-                url=url,
-                json=json,
-                timeout=timeout,
-            )
-            response.raise_for_status()
-            self.connected = True
-            self._retry_count = 0
-            return response.json()
-        except RequestException as e:
-            self._retry_count += 1
-            if self._retry_count >= self._max_retries:
-                self.connected = False
-            return None
-    
-    def check_connection(self) -> bool:
-        """Check if the backend is reachable."""
-        try:
-            response = self.session.get(
-                f"{self.base_url}/health",
-                timeout=2.0,
-            )
-            self.connected = response.status_code == 200
-            return self.connected
-        except RequestException:
-            self.connected = False
-            return False
-    
-    def get_state(self) -> Optional[SimulationState]:
-        """Fetch the current simulation state."""
-        data = self._request("GET", "/state")
-        if data:
-            self.last_state = SimulationState.from_api_response(data)
-            return self.last_state
-        return self.last_state  # Return cached state if request failed
-    
-    def advance_turn(self) -> bool:
-        """Advance the simulation by one step."""
-        result = self._request("POST", "/control/next_step")
-        return result is not None and result.get("success", False)
-    
-    def set_mode(self, mode: str) -> bool:
-        """Set the simulation mode ('manual' or 'auto')."""
-        result = self._request("POST", "/control/mode", json={"mode": mode})
-        return result is not None and result.get("success", False)
-    
-    def initialize(
-        self,
-        num_agents: int = 8,
-        world_width: int = 20,
-        world_height: int = 20,
-    ) -> bool:
-        """Initialize or reset the simulation."""
-        result = self._request("POST", "/control/initialize", json={
-            "num_agents": num_agents,
-            "world_width": world_width,
-            "world_height": world_height,
-        })
-        return result is not None and result.get("success", False)
-    
-    def get_status(self) -> Optional[dict]:
-        """Get simulation status."""
-        return self._request("GET", "/control/status")
-    
-    def get_agents(self) -> Optional[list[dict]]:
-        """Get all agents."""
-        result = self._request("GET", "/agents")
-        if result:
-            return result.get("agents", [])
-        return None
-    
-    def get_market_orders(self) -> Optional[list[dict]]:
-        """Get all market orders."""
-        result = self._request("GET", "/market/orders")
-        if result:
-            return result.get("orders", [])
-        return None
-    
-    def get_market_prices(self) -> Optional[dict]:
-        """Get market prices."""
-        return self._request("GET", "/market/prices")
-    
-    def wait_for_connection(self, timeout: float = 30.0) -> bool:
-        """Wait for backend connection with timeout."""
-        start = time.time()
-        while time.time() - start < timeout:
-            if self.check_connection():
-                return True
-            time.sleep(0.5)
-        return False
-    
-    def get_config(self) -> Optional[dict]:
-        """Get current simulation configuration."""
-        return self._request("GET", "/config")
-    
-    def update_config(self, config_data: dict) -> bool:
-        """Update simulation configuration."""
-        result = self._request("POST", "/config", json=config_data)
-        return result is not None and result.get("success", False)
-    
-    def reset_config(self) -> bool:
-        """Reset configuration to defaults."""
-        result = self._request("POST", "/config/reset")
-        return result is not None and result.get("success", False)
-

frontend/main.py

@@ -1,324 +0,0 @@
-"""Main Pygame application for the Village Simulation frontend."""
-
-import sys
-import pygame
-
-from frontend.client import SimulationClient, SimulationState
-from frontend.renderer.map_renderer import MapRenderer
-from frontend.renderer.agent_renderer import AgentRenderer
-from frontend.renderer.ui_renderer import UIRenderer
-from frontend.renderer.settings_renderer import SettingsRenderer
-from frontend.renderer.stats_renderer import StatsRenderer
-
-
-# Window configuration
-WINDOW_WIDTH = 1200
-WINDOW_HEIGHT = 800
-WINDOW_TITLE = "Village Economy Simulation"
-FPS = 30
-
-# Layout configuration
-TOP_PANEL_HEIGHT = 50
-RIGHT_PANEL_WIDTH = 200
-
-
-class VillageSimulationApp:
-    """Main application class for the Village Simulation frontend."""
-    
-    def __init__(self, server_url: str = "http://localhost:8000"):
-        # Initialize Pygame
-        pygame.init()
-        pygame.font.init()
-        
-        # Create window
-        self.screen = pygame.display.set_mode((WINDOW_WIDTH, WINDOW_HEIGHT))
-        pygame.display.set_caption(WINDOW_TITLE)
-        
-        # Clock for FPS control
-        self.clock = pygame.time.Clock()
-        
-        # Fonts
-        self.font = pygame.font.Font(None, 24)
-        
-        # Network client
-        self.client = SimulationClient(server_url)
-        
-        # Calculate map area
-        self.map_rect = pygame.Rect(
-            0,
-            TOP_PANEL_HEIGHT,
-            WINDOW_WIDTH - RIGHT_PANEL_WIDTH,
-            WINDOW_HEIGHT - TOP_PANEL_HEIGHT,
-        )
-        
-        # Initialize renderers
-        self.map_renderer = MapRenderer(self.screen, self.map_rect)
-        self.agent_renderer = AgentRenderer(self.screen, self.map_renderer, self.font)
-        self.ui_renderer = UIRenderer(self.screen, self.font)
-        self.settings_renderer = SettingsRenderer(self.screen)
-        self.stats_renderer = StatsRenderer(self.screen)
-        
-        # State
-        self.state: SimulationState | None = None
-        self.running = True
-        self.hovered_agent: dict | None = None
-        self._last_turn: int = -1  # Track turn changes for stats update
-        
-        # Polling interval (ms)
-        self.last_poll_time = 0
-        self.poll_interval = 100  # Poll every 100ms for smoother updates
-        
-        # Setup settings callbacks
-        self._setup_settings_callbacks()
-    
-    def _setup_settings_callbacks(self) -> None:
-        """Set up callbacks for the settings panel."""
-        # Override the apply and reset callbacks
-        original_apply = self.settings_renderer._apply_config
-        original_reset = self.settings_renderer._reset_config
-        
-        def apply_config():
-            config = self.settings_renderer.get_config()
-            if self.client.update_config(config):
-                # Restart simulation with new config
-                if self.client.initialize():
-                    self.state = self.client.get_state()
-                    self.settings_renderer.status_message = "Config applied & simulation restarted!"
-                    self.settings_renderer.status_color = (80, 180, 100)
-                else:
-                    self.settings_renderer.status_message = "Config saved but restart failed"
-                    self.settings_renderer.status_color = (200, 160, 80)
-            else:
-                self.settings_renderer.status_message = "Failed to apply config"
-                self.settings_renderer.status_color = (200, 80, 80)
-        
-        def reset_config():
-            if self.client.reset_config():
-                # Reload config from server
-                config = self.client.get_config()
-                if config:
-                    self.settings_renderer.set_config(config)
-                self.settings_renderer.status_message = "Config reset to defaults"
-                self.settings_renderer.status_color = (200, 160, 80)
-            else:
-                self.settings_renderer.status_message = "Failed to reset config"
-                self.settings_renderer.status_color = (200, 80, 80)
-        
-        self.settings_renderer._apply_config = apply_config
-        self.settings_renderer._reset_config = reset_config
-    
-    def _load_config(self) -> None:
-        """Load configuration from server into settings panel."""
-        config = self.client.get_config()
-        if config:
-            self.settings_renderer.set_config(config)
-    
-    def handle_events(self) -> None:
-        """Handle Pygame events."""
-        for event in pygame.event.get():
-            if event.type == pygame.QUIT:
-                self.running = False
-            
-            # Let stats panel handle events first if visible
-            if self.stats_renderer.handle_event(event):
-                continue
-            
-            # Let settings panel handle events first if visible
-            if self.settings_renderer.handle_event(event):
-                continue
-            
-            if event.type == pygame.KEYDOWN:
-                self._handle_keydown(event)
-            
-            elif event.type == pygame.MOUSEMOTION:
-                self._handle_mouse_motion(event)
-    
-    def _handle_keydown(self, event: pygame.event.Event) -> None:
-        """Handle keyboard input."""
-        if event.key == pygame.K_ESCAPE:
-            if self.stats_renderer.visible:
-                self.stats_renderer.toggle()
-            elif self.settings_renderer.visible:
-                self.settings_renderer.toggle()
-            else:
-                self.running = False
-        
-        elif event.key == pygame.K_SPACE:
-            # Advance one turn
-            if self.client.connected and not self.settings_renderer.visible and not self.stats_renderer.visible:
-                if self.client.advance_turn():
-                    # Immediately fetch new state
-                    self.state = self.client.get_state()
-        
-        elif event.key == pygame.K_r:
-            # Reset simulation
-            if self.client.connected and not self.settings_renderer.visible and not self.stats_renderer.visible:
-                if self.client.initialize():
-                    self.state = self.client.get_state()
-                    self.stats_renderer.clear_history()
-                    self._last_turn = -1
-        
-        elif event.key == pygame.K_m:
-            # Toggle mode
-            if self.client.connected and self.state and not self.settings_renderer.visible and not self.stats_renderer.visible:
-                new_mode = "auto" if self.state.mode == "manual" else "manual"
-                if self.client.set_mode(new_mode):
-                    self.state = self.client.get_state()
-        
-        elif event.key == pygame.K_g:
-            # Toggle statistics/graphs panel
-            if not self.settings_renderer.visible:
-                self.stats_renderer.toggle()
-        
-        elif event.key == pygame.K_s:
-            # Toggle settings panel
-            if not self.stats_renderer.visible:
-                if not self.settings_renderer.visible:
-                    self._load_config()
-                self.settings_renderer.toggle()
-    
-    def _handle_mouse_motion(self, event: pygame.event.Event) -> None:
-        """Handle mouse motion for agent hover detection."""
-        if not self.state or self.settings_renderer.visible:
-            self.hovered_agent = None
-            return
-        
-        mouse_pos = event.pos
-        self.hovered_agent = None
-        
-        # Check if mouse is in map area
-        if not self.map_rect.collidepoint(mouse_pos):
-            return
-        
-        # Check each agent
-        for agent in self.state.agents:
-            if not agent.get("is_alive", False):
-                continue
-            
-            pos = agent.get("position", {"x": 0, "y": 0})
-            screen_x, screen_y = self.map_renderer.grid_to_screen(pos["x"], pos["y"])
-            
-            # Check if mouse is near agent
-            dx = mouse_pos[0] - screen_x
-            dy = mouse_pos[1] - screen_y
-            distance = (dx * dx + dy * dy) ** 0.5
-            
-            cell_w, cell_h = self.map_renderer.get_cell_size()
-            agent_radius = min(cell_w, cell_h) / 2
-            
-            if distance < agent_radius + 5:
-                self.hovered_agent = agent
-                break
-    
-    def update(self) -> None:
-        """Update game state by polling the server."""
-        current_time = pygame.time.get_ticks()
-        
-        # Check if we need to poll
-        if current_time - self.last_poll_time >= self.poll_interval:
-            self.last_poll_time = current_time
-            
-            if not self.client.connected:
-                self.client.check_connection()
-            
-            if self.client.connected:
-                new_state = self.client.get_state()
-                if new_state:
-                    # Update map dimensions if changed
-                    if (
-                        new_state.world_width != self.map_renderer.world_width or
-                        new_state.world_height != self.map_renderer.world_height
-                    ):
-                        self.map_renderer.update_dimensions(
-                            new_state.world_width,
-                            new_state.world_height,
-                        )
-                    self.state = new_state
-                    
-                    # Update stats history when turn changes
-                    if new_state.turn != self._last_turn:
-                        self.stats_renderer.update_history(new_state)
-                        self._last_turn = new_state.turn
-    
-    def draw(self) -> None:
-        """Draw all elements."""
-        # Clear screen
-        self.screen.fill((30, 35, 45))
-        
-        if self.state:
-            # Draw map
-            self.map_renderer.draw(self.state)
-            
-            # Draw agents
-            self.agent_renderer.draw(self.state)
-            
-            # Draw UI
-            self.ui_renderer.draw(self.state)
-            
-            # Draw agent tooltip if hovering
-            if self.hovered_agent and not self.settings_renderer.visible:
-                mouse_pos = pygame.mouse.get_pos()
-                self.agent_renderer.draw_agent_tooltip(self.hovered_agent, mouse_pos)
-        
-        # Draw connection status overlay if disconnected
-        if not self.client.connected:
-            self.ui_renderer.draw_connection_status(self.client.connected)
-        
-        # Draw settings panel if visible
-        self.settings_renderer.draw()
-        
-        # Draw stats panel if visible
-        self.stats_renderer.draw(self.state)
-        
-        # Draw hints at bottom
-        if not self.settings_renderer.visible and not self.stats_renderer.visible:
-            hint_font = pygame.font.Font(None, 18)
-            hint = hint_font.render("S: Settings  |  G: Statistics & Graphs", True, (100, 100, 120))
-            self.screen.blit(hint, (5, self.screen.get_height() - 20))
-        
-        # Update display
-        pygame.display.flip()
-    
-    def run(self) -> None:
-        """Main game loop."""
-        print("Starting Village Simulation Frontend...")
-        print("Connecting to backend at http://localhost:8000...")
-        
-        # Try to connect initially
-        if not self.client.check_connection():
-            print("Backend not available. Will retry in the main loop.")
-        else:
-            print("Connected!")
-            self.state = self.client.get_state()
-        
-        print("\nControls:")
-        print("  SPACE - Advance turn")
-        print("  R     - Reset simulation")
-        print("  M     - Toggle auto/manual mode")
-        print("  S     - Open settings")
-        print("  G     - Open statistics & graphs")
-        print("  ESC   - Close panel / Quit")
-        print()
-        
-        while self.running:
-            self.handle_events()
-            self.update()
-            self.draw()
-            self.clock.tick(FPS)
-        
-        pygame.quit()
-
-
-def main():
-    """Entry point for the frontend application."""
-    # Get server URL from command line if provided
-    server_url = "http://localhost:8000"
-    if len(sys.argv) > 1:
-        server_url = sys.argv[1]
-    
-    app = VillageSimulationApp(server_url)
-    app.run()
-
-
-if __name__ == "__main__":
-    main()

frontend/renderer/__init__.py

@@ -1,9 +0,0 @@
-"""Renderer components for the Village Simulation frontend."""
-
-from .map_renderer import MapRenderer
-from .agent_renderer import AgentRenderer
-from .ui_renderer import UIRenderer
-from .settings_renderer import SettingsRenderer
-
-__all__ = ["MapRenderer", "AgentRenderer", "UIRenderer", "SettingsRenderer"]
-

frontend/renderer/agent_renderer.py

@@ -1,430 +0,0 @@
-"""Agent renderer for the Village Simulation."""
-
-import math
-import pygame
-from typing import TYPE_CHECKING
-
-if TYPE_CHECKING:
-    from frontend.client import SimulationState
-    from frontend.renderer.map_renderer import MapRenderer
-
-
-# Profession colors (villager is the default now)
-PROFESSION_COLORS = {
-    "villager": (100, 140, 180),  # Blue-gray for generic villager
-    "hunter": (180, 80, 80),      # Red
-    "gatherer": (80, 160, 80),    # Green
-    "woodcutter": (139, 90, 43),  # Brown
-    "crafter": (160, 120, 200),   # Purple
-}
-
-# Corpse color
-CORPSE_COLOR = (60, 60, 60)  # Dark gray
-
-# Status bar colors
-BAR_COLORS = {
-    "energy": (255, 220, 80),   # Yellow
-    "hunger": (220, 140, 80),   # Orange
-    "thirst": (80, 160, 220),   # Blue
-    "heat": (220, 80, 80),      # Red
-}
-
-# Action icons/symbols
-ACTION_SYMBOLS = {
-    "hunt": "🏹",
-    "gather": "🍇",
-    "chop_wood": "🪓",
-    "get_water": "💧",
-    "weave": "🧵",
-    "build_fire": "🔥",
-    "trade": "💰",
-    "rest": "💤",
-    "sleep": "😴",
-    "consume": "🍖",
-    "dead": "💀",
-}
-
-# Fallback ASCII symbols for systems without emoji support
-ACTION_LETTERS = {
-    "hunt": "H",
-    "gather": "G",
-    "chop_wood": "W",
-    "get_water": "~",
-    "weave": "C",
-    "build_fire": "F",
-    "trade": "$",
-    "rest": "R",
-    "sleep": "Z",
-    "consume": "E",
-    "dead": "X",
-}
-
-
-class AgentRenderer:
-    """Renders agents on the map with movement and action indicators."""
-    
-    def __init__(
-        self,
-        screen: pygame.Surface,
-        map_renderer: "MapRenderer",
-        font: pygame.font.Font,
-    ):
-        self.screen = screen
-        self.map_renderer = map_renderer
-        self.font = font
-        self.small_font = pygame.font.Font(None, 16)
-        self.action_font = pygame.font.Font(None, 20)
-        
-        # Animation state
-        self.animation_tick = 0
-    
-    def _get_agent_color(self, agent: dict) -> tuple[int, int, int]:
-        """Get the color for an agent based on state."""
-        # Corpses are dark gray
-        if agent.get("is_corpse", False) or not agent.get("is_alive", True):
-            return CORPSE_COLOR
-        
-        profession = agent.get("profession", "villager")
-        base_color = PROFESSION_COLORS.get(profession, (100, 140, 180))
-        
-        if not agent.get("can_act", True):
-            # Slightly dimmed for exhausted agents
-            return tuple(int(c * 0.7) for c in base_color)
-        
-        return base_color
-    
-    def _draw_status_bar(
-        self,
-        x: int,
-        y: int,
-        width: int,
-        height: int,
-        value: int,
-        max_value: int,
-        color: tuple[int, int, int],
-    ) -> None:
-        """Draw a single status bar."""
-        # Background
-        pygame.draw.rect(self.screen, (40, 40, 40), (x, y, width, height))
-        
-        # Fill
-        fill_width = int((value / max_value) * width) if max_value > 0 else 0
-        if fill_width > 0:
-            pygame.draw.rect(self.screen, color, (x, y, fill_width, height))
-        
-        # Border
-        pygame.draw.rect(self.screen, (80, 80, 80), (x, y, width, height), 1)
-    
-    def _draw_status_bars(self, agent: dict, center_x: int, center_y: int, size: int) -> None:
-        """Draw status bars below the agent."""
-        stats = agent.get("stats", {})
-        
-        bar_width = size + 10
-        bar_height = 3
-        bar_spacing = 4
-        start_y = center_y + size // 2 + 4
-        
-        bars = [
-            ("energy", stats.get("energy", 0), stats.get("max_energy", 100)),
-            ("hunger", stats.get("hunger", 0), stats.get("max_hunger", 100)),
-            ("thirst", stats.get("thirst", 0), stats.get("max_thirst", 50)),
-            ("heat", stats.get("heat", 0), stats.get("max_heat", 100)),
-        ]
-        
-        for i, (stat_name, value, max_value) in enumerate(bars):
-            bar_y = start_y + i * bar_spacing
-            self._draw_status_bar(
-                center_x - bar_width // 2,
-                bar_y,
-                bar_width,
-                bar_height,
-                value,
-                max_value,
-                BAR_COLORS[stat_name],
-            )
-    
-    def _draw_action_indicator(
-        self,
-        agent: dict,
-        center_x: int,
-        center_y: int,
-        agent_size: int,
-    ) -> None:
-        """Draw action indicator above the agent."""
-        current_action = agent.get("current_action", {})
-        action_type = current_action.get("action_type", "")
-        is_moving = current_action.get("is_moving", False)
-        message = current_action.get("message", "")
-        
-        if not action_type:
-            return
-        
-        # Get action symbol
-        symbol = ACTION_LETTERS.get(action_type, "?")
-        
-        # Draw action bubble above agent
-        bubble_y = center_y - agent_size // 2 - 20
-        
-        # Animate if moving
-        if is_moving:
-            # Bouncing animation
-            offset = int(3 * math.sin(self.animation_tick * 0.3))
-            bubble_y += offset
-        
-        # Draw bubble background
-        bubble_width = 22
-        bubble_height = 18
-        bubble_rect = pygame.Rect(
-            center_x - bubble_width // 2,
-            bubble_y - bubble_height // 2,
-            bubble_width,
-            bubble_height,
-        )
-        
-        # Color based on action success/failure
-        if "Failed" in message:
-            bg_color = (120, 60, 60)
-            border_color = (180, 80, 80)
-        elif is_moving:
-            bg_color = (60, 80, 120)
-            border_color = (100, 140, 200)
-        else:
-            bg_color = (50, 70, 50)
-            border_color = (80, 140, 80)
-        
-        pygame.draw.rect(self.screen, bg_color, bubble_rect, border_radius=4)
-        pygame.draw.rect(self.screen, border_color, bubble_rect, 1, border_radius=4)
-        
-        # Draw action letter
-        text = self.action_font.render(symbol, True, (255, 255, 255))
-        text_rect = text.get_rect(center=(center_x, bubble_y))
-        self.screen.blit(text, text_rect)
-        
-        # Draw movement trail if moving
-        if is_moving:
-            target_pos = current_action.get("target_position")
-            if target_pos:
-                target_x, target_y = self.map_renderer.grid_to_screen(
-                    target_pos.get("x", 0),
-                    target_pos.get("y", 0),
-                )
-                # Draw dotted line to target
-                self._draw_dotted_line(
-                    (center_x, center_y),
-                    (target_x, target_y),
-                    (100, 100, 100),
-                    4,
-                )
-    
-    def _draw_dotted_line(
-        self,
-        start: tuple[int, int],
-        end: tuple[int, int],
-        color: tuple[int, int, int],
-        dot_spacing: int = 5,
-    ) -> None:
-        """Draw a dotted line between two points."""
-        dx = end[0] - start[0]
-        dy = end[1] - start[1]
-        distance = max(1, int((dx ** 2 + dy ** 2) ** 0.5))
-        
-        for i in range(0, distance, dot_spacing * 2):
-            t = i / distance
-            x = int(start[0] + dx * t)
-            y = int(start[1] + dy * t)
-            pygame.draw.circle(self.screen, color, (x, y), 1)
-    
-    def _draw_last_action_result(
-        self,
-        agent: dict,
-        center_x: int,
-        center_y: int,
-        agent_size: int,
-    ) -> None:
-        """Draw the last action result as floating text."""
-        result = agent.get("last_action_result", "")
-        if not result:
-            return
-        
-        # Truncate long messages
-        if len(result) > 25:
-            result = result[:22] + "..."
-        
-        # Draw text below status bars
-        text_y = center_y + agent_size // 2 + 22
-        
-        text = self.small_font.render(result, True, (180, 180, 180))
-        text_rect = text.get_rect(center=(center_x, text_y))
-        
-        # Background for readability
-        bg_rect = text_rect.inflate(4, 2)
-        pygame.draw.rect(self.screen, (30, 30, 40, 180), bg_rect)
-        
-        self.screen.blit(text, text_rect)
-    
-    def draw(self, state: "SimulationState") -> None:
-        """Draw all agents (including corpses for one turn)."""
-        self.animation_tick += 1
-        
-        cell_w, cell_h = self.map_renderer.get_cell_size()
-        agent_size = min(cell_w, cell_h) - 8
-        agent_size = max(10, min(agent_size, 30))  # Clamp size
-        
-        for agent in state.agents:
-            is_corpse = agent.get("is_corpse", False)
-            is_alive = agent.get("is_alive", True)
-            
-            # Get screen position from agent's current position
-            pos = agent.get("position", {"x": 0, "y": 0})
-            screen_x, screen_y = self.map_renderer.grid_to_screen(pos["x"], pos["y"])
-            
-            if is_corpse:
-                # Draw corpse with death indicator
-                self._draw_corpse(agent, screen_x, screen_y, agent_size)
-                continue
-            
-            if not is_alive:
-                continue
-            
-            # Draw movement trail/line to target first (behind agent)
-            self._draw_action_indicator(agent, screen_x, screen_y, agent_size)
-            
-            # Draw agent circle
-            color = self._get_agent_color(agent)
-            pygame.draw.circle(self.screen, color, (screen_x, screen_y), agent_size // 2)
-            
-            # Draw border - animated if moving
-            current_action = agent.get("current_action", {})
-            is_moving = current_action.get("is_moving", False)
-            
-            if is_moving:
-                # Pulsing border when moving
-                pulse = int(127 + 127 * math.sin(self.animation_tick * 0.2))
-                border_color = (pulse, pulse, 255)
-            elif agent.get("can_act"):
-                border_color = (255, 255, 255)
-            else:
-                border_color = (100, 100, 100)
-            
-            pygame.draw.circle(self.screen, border_color, (screen_x, screen_y), agent_size // 2, 2)
-            
-            # Draw money indicator (small coin icon)
-            money = agent.get("money", 0)
-            if money > 0:
-                coin_x = screen_x + agent_size // 2 - 4
-                coin_y = screen_y - agent_size // 2 - 4
-                pygame.draw.circle(self.screen, (255, 215, 0), (coin_x, coin_y), 4)
-                pygame.draw.circle(self.screen, (200, 160, 0), (coin_x, coin_y), 4, 1)
-            
-            # Draw "V" for villager
-            text = self.small_font.render("V", True, (255, 255, 255))
-            text_rect = text.get_rect(center=(screen_x, screen_y))
-            self.screen.blit(text, text_rect)
-            
-            # Draw status bars
-            self._draw_status_bars(agent, screen_x, screen_y, agent_size)
-            
-            # Draw last action result
-            self._draw_last_action_result(agent, screen_x, screen_y, agent_size)
-    
-    def _draw_corpse(
-        self,
-        agent: dict,
-        center_x: int,
-        center_y: int,
-        agent_size: int,
-    ) -> None:
-        """Draw a corpse with death reason displayed."""
-        # Draw corpse circle (dark gray)
-        pygame.draw.circle(self.screen, CORPSE_COLOR, (center_x, center_y), agent_size // 2)
-        
-        # Draw red X border
-        pygame.draw.circle(self.screen, (150, 50, 50), (center_x, center_y), agent_size // 2, 2)
-        
-        # Draw skull symbol
-        text = self.action_font.render("X", True, (180, 80, 80))
-        text_rect = text.get_rect(center=(center_x, center_y))
-        self.screen.blit(text, text_rect)
-        
-        # Draw death reason above corpse
-        death_reason = agent.get("death_reason", "unknown")
-        name = agent.get("name", "Unknown")
-        
-        # Death indicator bubble
-        bubble_y = center_y - agent_size // 2 - 20
-        bubble_text = f"💀 {death_reason}"
-        
-        text = self.small_font.render(bubble_text, True, (255, 100, 100))
-        text_rect = text.get_rect(center=(center_x, bubble_y))
-        
-        # Background for readability
-        bg_rect = text_rect.inflate(8, 4)
-        pygame.draw.rect(self.screen, (40, 20, 20), bg_rect, border_radius=3)
-        pygame.draw.rect(self.screen, (120, 50, 50), bg_rect, 1, border_radius=3)
-        
-        self.screen.blit(text, text_rect)
-        
-        # Draw name below
-        name_y = center_y + agent_size // 2 + 8
-        name_text = self.small_font.render(name, True, (150, 150, 150))
-        name_rect = name_text.get_rect(center=(center_x, name_y))
-        self.screen.blit(name_text, name_rect)
-    
-    def draw_agent_tooltip(self, agent: dict, mouse_pos: tuple[int, int]) -> None:
-        """Draw a tooltip for an agent when hovered."""
-        # Build tooltip text
-        lines = [
-            agent.get("name", "Unknown"),
-            f"Profession: {agent.get('profession', '?').capitalize()}",
-            f"Money: {agent.get('money', 0)} coins",
-            "",
-        ]
-        
-        # Current action
-        current_action = agent.get("current_action", {})
-        action_type = current_action.get("action_type", "")
-        if action_type:
-            action_msg = current_action.get("message", action_type)
-            lines.append(f"Action: {action_msg[:40]}")
-            if current_action.get("is_moving"):
-                lines.append("  (moving to location)")
-            lines.append("")
-        
-        lines.append("Stats:")
-        stats = agent.get("stats", {})
-        lines.append(f"  Energy: {stats.get('energy', 0)}/{stats.get('max_energy', 100)}")
-        lines.append(f"  Hunger: {stats.get('hunger', 0)}/{stats.get('max_hunger', 100)}")
-        lines.append(f"  Thirst: {stats.get('thirst', 0)}/{stats.get('max_thirst', 50)}")
-        lines.append(f"  Heat: {stats.get('heat', 0)}/{stats.get('max_heat', 100)}")
-        
-        inventory = agent.get("inventory", [])
-        if inventory:
-            lines.append("")
-            lines.append("Inventory:")
-            for item in inventory[:5]:
-                lines.append(f"  {item.get('type', '?')}: {item.get('quantity', 0)}")
-        
-        # Last action result
-        last_result = agent.get("last_action_result", "")
-        if last_result:
-            lines.append("")
-            lines.append(f"Last: {last_result[:35]}")
-        
-        # Calculate tooltip size
-        line_height = 16
-        max_width = max(self.small_font.size(line)[0] for line in lines) + 20
-        height = len(lines) * line_height + 10
-        
-        # Position tooltip near mouse but not off screen
-        x = min(mouse_pos[0] + 15, self.screen.get_width() - max_width - 5)
-        y = min(mouse_pos[1] + 15, self.screen.get_height() - height - 5)
-        
-        # Draw background
-        tooltip_rect = pygame.Rect(x, y, max_width, height)
-        pygame.draw.rect(self.screen, (30, 30, 40), tooltip_rect)
-        pygame.draw.rect(self.screen, (100, 100, 120), tooltip_rect, 1)
-        
-        # Draw text
-        for i, line in enumerate(lines):
-            text = self.small_font.render(line, True, (220, 220, 220))
-            self.screen.blit(text, (x + 10, y + 5 + i * line_height))

frontend/renderer/map_renderer.py

@@ -1,146 +0,0 @@
-"""Map renderer for the Village Simulation."""
-
-import pygame
-from typing import TYPE_CHECKING
-
-if TYPE_CHECKING:
-    from frontend.client import SimulationState
-
-
-# Color palette
-class Colors:
-    # Background colors
-    DAY_BG = (180, 200, 160)       # Soft green for day
-    NIGHT_BG = (40, 45, 60)        # Dark blue for night
-    GRID_LINE = (120, 140, 110)   # Subtle grid lines
-    GRID_LINE_NIGHT = (60, 65, 80)
-    
-    # Terrain features (for visual variety)
-    GRASS_LIGHT = (160, 190, 140)
-    GRASS_DARK = (140, 170, 120)
-    WATER_SPOT = (100, 140, 180)
-
-
-class MapRenderer:
-    """Renders the map/terrain background."""
-    
-    def __init__(
-        self,
-        screen: pygame.Surface,
-        map_rect: pygame.Rect,
-        world_width: int = 20,
-        world_height: int = 20,
-    ):
-        self.screen = screen
-        self.map_rect = map_rect
-        self.world_width = world_width
-        self.world_height = world_height
-        self._cell_width = map_rect.width / world_width
-        self._cell_height = map_rect.height / world_height
-        
-        # Pre-generate some terrain variation
-        self._terrain_cache = self._generate_terrain()
-    
-    def _generate_terrain(self) -> list[list[int]]:
-        """Generate simple terrain variation (0 = light, 1 = dark, 2 = water)."""
-        import random
-        terrain = []
-        for y in range(self.world_height):
-            row = []
-            for x in range(self.world_width):
-                # Simple pattern: mostly grass with occasional water spots
-                if random.random() < 0.05:
-                    row.append(2)  # Water spot
-                elif (x + y) % 3 == 0:
-                    row.append(1)  # Dark grass
-                else:
-                    row.append(0)  # Light grass
-            terrain.append(row)
-        return terrain
-    
-    def update_dimensions(self, world_width: int, world_height: int) -> None:
-        """Update world dimensions and recalculate cell sizes."""
-        if world_width != self.world_width or world_height != self.world_height:
-            self.world_width = world_width
-            self.world_height = world_height
-            self._cell_width = self.map_rect.width / world_width
-            self._cell_height = self.map_rect.height / world_height
-            self._terrain_cache = self._generate_terrain()
-    
-    def grid_to_screen(self, grid_x: int, grid_y: int) -> tuple[int, int]:
-        """Convert grid coordinates to screen coordinates (center of cell)."""
-        screen_x = self.map_rect.left + (grid_x + 0.5) * self._cell_width
-        screen_y = self.map_rect.top + (grid_y + 0.5) * self._cell_height
-        return int(screen_x), int(screen_y)
-    
-    def get_cell_size(self) -> tuple[int, int]:
-        """Get the size of a single cell."""
-        return int(self._cell_width), int(self._cell_height)
-    
-    def draw(self, state: "SimulationState") -> None:
-        """Draw the map background."""
-        is_night = state.time_of_day == "night"
-        
-        # Fill background
-        bg_color = Colors.NIGHT_BG if is_night else Colors.DAY_BG
-        pygame.draw.rect(self.screen, bg_color, self.map_rect)
-        
-        # Draw terrain cells
-        for y in range(self.world_height):
-            for x in range(self.world_width):
-                cell_rect = pygame.Rect(
-                    self.map_rect.left + x * self._cell_width,
-                    self.map_rect.top + y * self._cell_height,
-                    self._cell_width + 1,  # +1 to avoid gaps
-                    self._cell_height + 1,
-                )
-                
-                terrain_type = self._terrain_cache[y][x]
-                
-                if is_night:
-                    # Darker colors at night
-                    if terrain_type == 2:
-                        color = (60, 80, 110)
-                    elif terrain_type == 1:
-                        color = (35, 40, 55)
-                    else:
-                        color = (45, 50, 65)
-                else:
-                    if terrain_type == 2:
-                        color = Colors.WATER_SPOT
-                    elif terrain_type == 1:
-                        color = Colors.GRASS_DARK
-                    else:
-                        color = Colors.GRASS_LIGHT
-                
-                pygame.draw.rect(self.screen, color, cell_rect)
-        
-        # Draw grid lines
-        grid_color = Colors.GRID_LINE_NIGHT if is_night else Colors.GRID_LINE
-        
-        # Vertical lines
-        for x in range(self.world_width + 1):
-            start_x = self.map_rect.left + x * self._cell_width
-            pygame.draw.line(
-                self.screen,
-                grid_color,
-                (start_x, self.map_rect.top),
-                (start_x, self.map_rect.bottom),
-                1,
-            )
-        
-        # Horizontal lines
-        for y in range(self.world_height + 1):
-            start_y = self.map_rect.top + y * self._cell_height
-            pygame.draw.line(
-                self.screen,
-                grid_color,
-                (self.map_rect.left, start_y),
-                (self.map_rect.right, start_y),
-                1,
-            )
-        
-        # Draw border
-        border_color = (80, 90, 70) if not is_night else (80, 85, 100)
-        pygame.draw.rect(self.screen, border_color, self.map_rect, 2)
-

frontend/renderer/settings_renderer.py

@@ -1,448 +0,0 @@
-"""Settings UI renderer with sliders for the Village Simulation."""
-
-import pygame
-from dataclasses import dataclass
-from typing import Optional, Callable, Any
-
-
-class Colors:
-    """Color palette for settings UI."""
-    BG = (25, 28, 35)
-    PANEL_BG = (35, 40, 50)
-    PANEL_BORDER = (70, 80, 95)
-    TEXT_PRIMARY = (230, 230, 235)
-    TEXT_SECONDARY = (160, 165, 175)
-    TEXT_HIGHLIGHT = (100, 180, 255)
-    SLIDER_BG = (50, 55, 65)
-    SLIDER_FILL = (80, 140, 200)
-    SLIDER_HANDLE = (220, 220, 230)
-    BUTTON_BG = (60, 100, 160)
-    BUTTON_HOVER = (80, 120, 180)
-    BUTTON_TEXT = (255, 255, 255)
-    SUCCESS = (80, 180, 100)
-    WARNING = (200, 160, 80)
-
-
-@dataclass
-class SliderConfig:
-    """Configuration for a slider widget."""
-    name: str
-    key: str  # Dot-separated path like "agent_stats.max_energy"
-    min_val: float
-    max_val: float
-    step: float = 1.0
-    is_int: bool = True
-    description: str = ""
-
-
-# Define all configurable parameters with sliders
-SLIDER_CONFIGS = [
-    # Agent Stats Section
-    SliderConfig("Max Energy", "agent_stats.max_energy", 50, 200, 10, True, "Maximum energy capacity"),
-    SliderConfig("Max Hunger", "agent_stats.max_hunger", 50, 200, 10, True, "Maximum hunger capacity"),
-    SliderConfig("Max Thirst", "agent_stats.max_thirst", 25, 100, 5, True, "Maximum thirst capacity"),
-    SliderConfig("Max Heat", "agent_stats.max_heat", 50, 200, 10, True, "Maximum heat capacity"),
-    SliderConfig("Energy Decay", "agent_stats.energy_decay", 1, 10, 1, True, "Energy lost per turn"),
-    SliderConfig("Hunger Decay", "agent_stats.hunger_decay", 1, 10, 1, True, "Hunger lost per turn"),
-    SliderConfig("Thirst Decay", "agent_stats.thirst_decay", 1, 10, 1, True, "Thirst lost per turn"),
-    SliderConfig("Heat Decay", "agent_stats.heat_decay", 1, 10, 1, True, "Heat lost per turn"),
-    SliderConfig("Critical %", "agent_stats.critical_threshold", 0.1, 0.5, 0.05, False, "Threshold for survival mode"),
-    
-    # World Section
-    SliderConfig("World Width", "world.width", 10, 50, 5, True, "World grid width"),
-    SliderConfig("World Height", "world.height", 10, 50, 5, True, "World grid height"),
-    SliderConfig("Initial Agents", "world.initial_agents", 2, 20, 1, True, "Starting agent count"),
-    SliderConfig("Day Steps", "world.day_steps", 5, 20, 1, True, "Steps per day"),
-    SliderConfig("Inventory Slots", "world.inventory_slots", 5, 20, 1, True, "Agent inventory size"),
-    SliderConfig("Starting Money", "world.starting_money", 50, 500, 50, True, "Initial coins per agent"),
-    
-    # Actions Section
-    SliderConfig("Hunt Energy Cost", "actions.hunt_energy", -30, -5, 5, True, "Energy spent hunting"),
-    SliderConfig("Gather Energy Cost", "actions.gather_energy", -20, -1, 1, True, "Energy spent gathering"),
-    SliderConfig("Hunt Success %", "actions.hunt_success", 0.3, 1.0, 0.1, False, "Hunting success chance"),
-    SliderConfig("Sleep Restore", "actions.sleep_energy", 30, 100, 10, True, "Energy restored by sleep"),
-    SliderConfig("Rest Restore", "actions.rest_energy", 5, 30, 5, True, "Energy restored by rest"),
-    
-    # Resources Section
-    SliderConfig("Meat Decay", "resources.meat_decay", 2, 20, 1, True, "Turns until meat spoils"),
-    SliderConfig("Berries Decay", "resources.berries_decay", 10, 50, 5, True, "Turns until berries spoil"),
-    SliderConfig("Meat Hunger +", "resources.meat_hunger", 10, 60, 5, True, "Hunger restored by meat"),
-    SliderConfig("Water Thirst +", "resources.water_thirst", 20, 60, 5, True, "Thirst restored by water"),
-    
-    # Market Section
-    SliderConfig("Discount Turns", "market.turns_before_discount", 1, 10, 1, True, "Turns before price drop"),
-    SliderConfig("Discount Rate %", "market.discount_rate", 0.05, 0.30, 0.05, False, "Price reduction per period"),
-    
-    # Simulation Section
-    SliderConfig("Auto Step (s)", "auto_step_interval", 0.2, 3.0, 0.2, False, "Seconds between auto steps"),
-]
-
-
-class Slider:
-    """A slider widget for adjusting numeric values."""
-    
-    def __init__(
-        self,
-        rect: pygame.Rect,
-        config: SliderConfig,
-        font: pygame.font.Font,
-        small_font: pygame.font.Font,
-    ):
-        self.rect = rect
-        self.config = config
-        self.font = font
-        self.small_font = small_font
-        self.value = config.min_val
-        self.dragging = False
-        self.hovered = False
-    
-    def set_value(self, value: float) -> None:
-        """Set the slider value."""
-        self.value = max(self.config.min_val, min(self.config.max_val, value))
-        if self.config.is_int:
-            self.value = int(round(self.value))
-    
-    def get_value(self) -> Any:
-        """Get the current value."""
-        return int(self.value) if self.config.is_int else round(self.value, 2)
-    
-    def handle_event(self, event: pygame.event.Event) -> bool:
-        """Handle input events. Returns True if value changed."""
-        if event.type == pygame.MOUSEBUTTONDOWN:
-            if self._slider_area().collidepoint(event.pos):
-                self.dragging = True
-                return self._update_from_mouse(event.pos[0])
-        
-        elif event.type == pygame.MOUSEBUTTONUP:
-            self.dragging = False
-        
-        elif event.type == pygame.MOUSEMOTION:
-            self.hovered = self.rect.collidepoint(event.pos)
-            if self.dragging:
-                return self._update_from_mouse(event.pos[0])
-        
-        return False
-    
-    def _slider_area(self) -> pygame.Rect:
-        """Get the actual slider track area."""
-        return pygame.Rect(
-            self.rect.x + 120,  # Leave space for label
-            self.rect.y + 15,
-            self.rect.width - 180,  # Leave space for value display
-            20,
-        )
-    
-    def _update_from_mouse(self, mouse_x: int) -> bool:
-        """Update value based on mouse position."""
-        slider_area = self._slider_area()
-        
-        # Calculate position as 0-1
-        rel_x = mouse_x - slider_area.x
-        ratio = max(0, min(1, rel_x / slider_area.width))
-        
-        # Calculate value
-        range_val = self.config.max_val - self.config.min_val
-        new_value = self.config.min_val + ratio * range_val
-        
-        # Apply step
-        if self.config.step > 0:
-            new_value = round(new_value / self.config.step) * self.config.step
-        
-        old_value = self.value
-        self.set_value(new_value)
-        return abs(old_value - self.value) > 0.001
-    
-    def draw(self, screen: pygame.Surface) -> None:
-        """Draw the slider."""
-        # Background
-        if self.hovered:
-            pygame.draw.rect(screen, (45, 50, 60), self.rect)
-        
-        # Label
-        label = self.small_font.render(self.config.name, True, Colors.TEXT_PRIMARY)
-        screen.blit(label, (self.rect.x + 5, self.rect.y + 5))
-        
-        # Slider track
-        slider_area = self._slider_area()
-        pygame.draw.rect(screen, Colors.SLIDER_BG, slider_area, border_radius=3)
-        
-        # Slider fill
-        ratio = (self.value - self.config.min_val) / (self.config.max_val - self.config.min_val)
-        fill_width = int(ratio * slider_area.width)
-        fill_rect = pygame.Rect(slider_area.x, slider_area.y, fill_width, slider_area.height)
-        pygame.draw.rect(screen, Colors.SLIDER_FILL, fill_rect, border_radius=3)
-        
-        # Handle
-        handle_x = slider_area.x + fill_width
-        handle_rect = pygame.Rect(handle_x - 4, slider_area.y - 2, 8, slider_area.height + 4)
-        pygame.draw.rect(screen, Colors.SLIDER_HANDLE, handle_rect, border_radius=2)
-        
-        # Value display
-        value_str = str(self.get_value())
-        value_text = self.small_font.render(value_str, True, Colors.TEXT_HIGHLIGHT)
-        value_x = self.rect.right - 50
-        screen.blit(value_text, (value_x, self.rect.y + 5))
-        
-        # Description on hover
-        if self.hovered and self.config.description:
-            desc = self.small_font.render(self.config.description, True, Colors.TEXT_SECONDARY)
-            screen.blit(desc, (self.rect.x + 5, self.rect.y + 25))
-
-
-class Button:
-    """A simple button widget."""
-    
-    def __init__(
-        self,
-        rect: pygame.Rect,
-        text: str,
-        font: pygame.font.Font,
-        callback: Optional[Callable] = None,
-        color: tuple = Colors.BUTTON_BG,
-    ):
-        self.rect = rect
-        self.text = text
-        self.font = font
-        self.callback = callback
-        self.color = color
-        self.hovered = False
-    
-    def handle_event(self, event: pygame.event.Event) -> bool:
-        """Handle input events. Returns True if clicked."""
-        if event.type == pygame.MOUSEMOTION:
-            self.hovered = self.rect.collidepoint(event.pos)
-        
-        elif event.type == pygame.MOUSEBUTTONDOWN:
-            if self.rect.collidepoint(event.pos):
-                if self.callback:
-                    self.callback()
-                return True
-        
-        return False
-    
-    def draw(self, screen: pygame.Surface) -> None:
-        """Draw the button."""
-        color = Colors.BUTTON_HOVER if self.hovered else self.color
-        pygame.draw.rect(screen, color, self.rect, border_radius=5)
-        pygame.draw.rect(screen, Colors.PANEL_BORDER, self.rect, 1, border_radius=5)
-        
-        text = self.font.render(self.text, True, Colors.BUTTON_TEXT)
-        text_rect = text.get_rect(center=self.rect.center)
-        screen.blit(text, text_rect)
-
-
-class SettingsRenderer:
-    """Renders the settings UI panel with sliders."""
-    
-    def __init__(self, screen: pygame.Surface):
-        self.screen = screen
-        self.font = pygame.font.Font(None, 24)
-        self.small_font = pygame.font.Font(None, 18)
-        self.title_font = pygame.font.Font(None, 32)
-        
-        self.visible = False
-        self.scroll_offset = 0
-        self.max_scroll = 0
-        
-        # Create sliders
-        self.sliders: list[Slider] = []
-        self.buttons: list[Button] = []
-        self.config_data: dict = {}
-        
-        self._create_widgets()
-        self.status_message = ""
-        self.status_color = Colors.TEXT_SECONDARY
-    
-    def _create_widgets(self) -> None:
-        """Create slider widgets."""
-        panel_width = 400
-        slider_height = 45
-        start_y = 80
-        
-        panel_x = (self.screen.get_width() - panel_width) // 2
-        
-        for i, config in enumerate(SLIDER_CONFIGS):
-            rect = pygame.Rect(
-                panel_x + 10,
-                start_y + i * slider_height,
-                panel_width - 20,
-                slider_height,
-            )
-            slider = Slider(rect, config, self.font, self.small_font)
-            self.sliders.append(slider)
-        
-        # Calculate max scroll
-        total_height = len(SLIDER_CONFIGS) * slider_height + 150
-        visible_height = self.screen.get_height() - 150
-        self.max_scroll = max(0, total_height - visible_height)
-        
-        # Create buttons at the bottom
-        button_y = self.screen.get_height() - 60
-        button_width = 100
-        button_height = 35
-        
-        buttons_data = [
-            ("Apply & Restart", self._apply_config, Colors.SUCCESS),
-            ("Reset Defaults", self._reset_config, Colors.WARNING),
-            ("Close", self.toggle, Colors.PANEL_BORDER),
-        ]
-        
-        total_button_width = len(buttons_data) * button_width + (len(buttons_data) - 1) * 10
-        start_x = (self.screen.get_width() - total_button_width) // 2
-        
-        for i, (text, callback, color) in enumerate(buttons_data):
-            rect = pygame.Rect(
-                start_x + i * (button_width + 10),
-                button_y,
-                button_width,
-                button_height,
-            )
-            self.buttons.append(Button(rect, text, self.small_font, callback, color))
-    
-    def toggle(self) -> None:
-        """Toggle settings visibility."""
-        self.visible = not self.visible
-        if self.visible:
-            self.scroll_offset = 0
-    
-    def set_config(self, config_data: dict) -> None:
-        """Set slider values from config data."""
-        self.config_data = config_data
-        
-        for slider in self.sliders:
-            value = self._get_nested_value(config_data, slider.config.key)
-            if value is not None:
-                slider.set_value(value)
-    
-    def get_config(self) -> dict:
-        """Get current config from slider values."""
-        result = {}
-        
-        for slider in self.sliders:
-            self._set_nested_value(result, slider.config.key, slider.get_value())
-        
-        return result
-    
-    def _get_nested_value(self, data: dict, key: str) -> Any:
-        """Get a value from nested dict using dot notation."""
-        parts = key.split(".")
-        current = data
-        for part in parts:
-            if isinstance(current, dict) and part in current:
-                current = current[part]
-            else:
-                return None
-        return current
-    
-    def _set_nested_value(self, data: dict, key: str, value: Any) -> None:
-        """Set a value in nested dict using dot notation."""
-        parts = key.split(".")
-        current = data
-        for part in parts[:-1]:
-            if part not in current:
-                current[part] = {}
-            current = current[part]
-        current[parts[-1]] = value
-    
-    def _apply_config(self) -> None:
-        """Apply configuration callback (to be set externally)."""
-        self.status_message = "Config applied - restart to see changes"
-        self.status_color = Colors.SUCCESS
-    
-    def _reset_config(self) -> None:
-        """Reset configuration callback (to be set externally)."""
-        self.status_message = "Config reset to defaults"
-        self.status_color = Colors.WARNING
-    
-    def handle_event(self, event: pygame.event.Event) -> bool:
-        """Handle input events. Returns True if event was consumed."""
-        if not self.visible:
-            return False
-        
-        # Handle scrolling
-        if event.type == pygame.MOUSEWHEEL:
-            self.scroll_offset -= event.y * 30
-            self.scroll_offset = max(0, min(self.max_scroll, self.scroll_offset))
-            return True
-        
-        # Handle sliders
-        for slider in self.sliders:
-            # Adjust slider position for scroll
-            original_y = slider.rect.y
-            slider.rect.y -= self.scroll_offset
-            
-            if slider.handle_event(event):
-                slider.rect.y = original_y
-                return True
-            
-            slider.rect.y = original_y
-        
-        # Handle buttons
-        for button in self.buttons:
-            if button.handle_event(event):
-                return True
-        
-        # Consume all clicks when settings are visible
-        if event.type == pygame.MOUSEBUTTONDOWN:
-            return True
-        
-        return False
-    
-    def draw(self) -> None:
-        """Draw the settings panel."""
-        if not self.visible:
-            return
-        
-        # Dim background
-        overlay = pygame.Surface(self.screen.get_size(), pygame.SRCALPHA)
-        overlay.fill((0, 0, 0, 200))
-        self.screen.blit(overlay, (0, 0))
-        
-        # Panel background
-        panel_width = 420
-        panel_height = self.screen.get_height() - 40
-        panel_x = (self.screen.get_width() - panel_width) // 2
-        panel_rect = pygame.Rect(panel_x, 20, panel_width, panel_height)
-        pygame.draw.rect(self.screen, Colors.PANEL_BG, panel_rect, border_radius=10)
-        pygame.draw.rect(self.screen, Colors.PANEL_BORDER, panel_rect, 2, border_radius=10)
-        
-        # Title
-        title = self.title_font.render("Simulation Settings", True, Colors.TEXT_PRIMARY)
-        title_rect = title.get_rect(centerx=self.screen.get_width() // 2, y=35)
-        self.screen.blit(title, title_rect)
-        
-        # Create clipping region for scrollable area
-        clip_rect = pygame.Rect(panel_x, 70, panel_width, panel_height - 130)
-        
-        # Draw sliders with scroll offset
-        for slider in self.sliders:
-            # Adjust position for scroll
-            adjusted_rect = slider.rect.copy()
-            adjusted_rect.y -= self.scroll_offset
-            
-            # Only draw if visible
-            if clip_rect.colliderect(adjusted_rect):
-                # Temporarily move slider for drawing
-                original_y = slider.rect.y
-                slider.rect.y = adjusted_rect.y
-                slider.draw(self.screen)
-                slider.rect.y = original_y
-        
-        # Draw scroll indicator
-        if self.max_scroll > 0:
-            scroll_ratio = self.scroll_offset / self.max_scroll
-            scroll_height = max(30, int((clip_rect.height / (clip_rect.height + self.max_scroll)) * clip_rect.height))
-            scroll_y = clip_rect.y + int(scroll_ratio * (clip_rect.height - scroll_height))
-            scroll_rect = pygame.Rect(panel_rect.right - 8, scroll_y, 4, scroll_height)
-            pygame.draw.rect(self.screen, Colors.SLIDER_FILL, scroll_rect, border_radius=2)
-        
-        # Draw buttons
-        for button in self.buttons:
-            button.draw(self.screen)
-        
-        # Status message
-        if self.status_message:
-            status = self.small_font.render(self.status_message, True, self.status_color)
-            status_rect = status.get_rect(centerx=self.screen.get_width() // 2, y=self.screen.get_height() - 90)
-            self.screen.blit(status, status_rect)
-

frontend/renderer/stats_renderer.py

@@ -1,770 +0,0 @@
-"""Real-time statistics and charts renderer for the Village Simulation.
-
-Uses matplotlib to render charts to pygame surfaces for a seamless visualization experience.
-"""
-
-import io
-from dataclasses import dataclass, field
-from collections import deque
-from typing import TYPE_CHECKING, Optional
-
-import pygame
-import matplotlib
-matplotlib.use('Agg')  # Use non-interactive backend for pygame integration
-import matplotlib.pyplot as plt
-import matplotlib.ticker as mticker
-from matplotlib.figure import Figure
-import numpy as np
-
-if TYPE_CHECKING:
-    from frontend.client import SimulationState
-
-
-# Color scheme - dark cyberpunk inspired
-class ChartColors:
-    """Color palette for charts - dark theme with neon accents."""
-    BG = '#1a1d26'
-    PANEL = '#252a38'
-    GRID = '#2f3545'
-    TEXT = '#e0e0e8'
-    TEXT_DIM = '#7a7e8c'
-    
-    # Neon accents for data series
-    CYAN = '#00d4ff'
-    MAGENTA = '#ff0099'
-    LIME = '#39ff14'
-    ORANGE = '#ff6600'
-    PURPLE = '#9d4edd'
-    YELLOW = '#ffcc00'
-    TEAL = '#00ffa3'
-    PINK = '#ff1493'
-    
-    # Series colors for different resources/categories
-    SERIES = [CYAN, MAGENTA, LIME, ORANGE, PURPLE, YELLOW, TEAL, PINK]
-
-
-class UIColors:
-    """Color palette for pygame UI elements."""
-    BG = (26, 29, 38)
-    PANEL_BG = (37, 42, 56)
-    PANEL_BORDER = (70, 80, 100)
-    TEXT_PRIMARY = (224, 224, 232)
-    TEXT_SECONDARY = (122, 126, 140)
-    TEXT_HIGHLIGHT = (0, 212, 255)
-    TAB_ACTIVE = (0, 212, 255)
-    TAB_INACTIVE = (55, 60, 75)
-    TAB_HOVER = (75, 85, 110)
-
-
-@dataclass
-class HistoryData:
-    """Stores historical simulation data for charting."""
-    max_history: int = 200
-    
-    # Time series data
-    turns: deque = field(default_factory=lambda: deque(maxlen=200))
-    population: deque = field(default_factory=lambda: deque(maxlen=200))
-    deaths_cumulative: deque = field(default_factory=lambda: deque(maxlen=200))
-    
-    # Money/Wealth data
-    total_money: deque = field(default_factory=lambda: deque(maxlen=200))
-    avg_wealth: deque = field(default_factory=lambda: deque(maxlen=200))
-    gini_coefficient: deque = field(default_factory=lambda: deque(maxlen=200))
-    
-    # Price history per resource
-    prices: dict = field(default_factory=dict)  # resource -> deque of prices
-    
-    # Trade statistics
-    trade_volume: deque = field(default_factory=lambda: deque(maxlen=200))
-    
-    # Profession counts over time
-    professions: dict = field(default_factory=dict)  # profession -> deque of counts
-    
-    def clear(self) -> None:
-        """Clear all history data."""
-        self.turns.clear()
-        self.population.clear()
-        self.deaths_cumulative.clear()
-        self.total_money.clear()
-        self.avg_wealth.clear()
-        self.gini_coefficient.clear()
-        self.prices.clear()
-        self.trade_volume.clear()
-        self.professions.clear()
-    
-    def update(self, state: "SimulationState") -> None:
-        """Update history with new state data."""
-        turn = state.turn
-        
-        # Avoid duplicate entries for the same turn
-        if self.turns and self.turns[-1] == turn:
-            return
-        
-        self.turns.append(turn)
-        
-        # Population
-        living = len([a for a in state.agents if a.get("is_alive", False)])
-        self.population.append(living)
-        self.deaths_cumulative.append(state.statistics.get("total_agents_died", 0))
-        
-        # Wealth data
-        stats = state.statistics
-        self.total_money.append(stats.get("total_money_in_circulation", 0))
-        self.avg_wealth.append(stats.get("avg_money", 0))
-        self.gini_coefficient.append(stats.get("gini_coefficient", 0))
-        
-        # Price history from market
-        for resource, data in state.market_prices.items():
-            if resource not in self.prices:
-                self.prices[resource] = deque(maxlen=self.max_history)
-            
-            # Track lowest price (current market rate)
-            lowest = data.get("lowest_price")
-            avg = data.get("avg_sale_price")
-            # Use lowest price if available, else avg sale price
-            price = lowest if lowest is not None else avg
-            self.prices[resource].append(price)
-        
-        # Trade volume (from recent trades in market orders)
-        trades = len(state.market_orders)  # Active orders as proxy
-        self.trade_volume.append(trades)
-        
-        # Profession distribution
-        professions = stats.get("professions", {})
-        for prof, count in professions.items():
-            if prof not in self.professions:
-                self.professions[prof] = deque(maxlen=self.max_history)
-            self.professions[prof].append(count)
-        
-        # Pad missing professions with 0
-        for prof in self.professions:
-            if prof not in professions:
-                self.professions[prof].append(0)
-
-
-class ChartRenderer:
-    """Renders matplotlib charts to pygame surfaces."""
-    
-    def __init__(self, width: int, height: int):
-        self.width = width
-        self.height = height
-        self.dpi = 100
-        
-        # Configure matplotlib style
-        plt.style.use('dark_background')
-        plt.rcParams.update({
-            'figure.facecolor': ChartColors.BG,
-            'axes.facecolor': ChartColors.PANEL,
-            'axes.edgecolor': ChartColors.GRID,
-            'axes.labelcolor': ChartColors.TEXT,
-            'text.color': ChartColors.TEXT,
-            'xtick.color': ChartColors.TEXT_DIM,
-            'ytick.color': ChartColors.TEXT_DIM,
-            'grid.color': ChartColors.GRID,
-            'grid.alpha': 0.3,
-            'legend.facecolor': ChartColors.PANEL,
-            'legend.edgecolor': ChartColors.GRID,
-            'font.size': 9,
-            'axes.titlesize': 11,
-            'axes.titleweight': 'bold',
-        })
-    
-    def _fig_to_surface(self, fig: Figure) -> pygame.Surface:
-        """Convert a matplotlib figure to a pygame surface."""
-        buf = io.BytesIO()
-        fig.savefig(buf, format='png', dpi=self.dpi, 
-                    facecolor=ChartColors.BG, edgecolor='none',
-                    bbox_inches='tight', pad_inches=0.1)
-        buf.seek(0)
-        
-        surface = pygame.image.load(buf, 'png')
-        buf.close()
-        plt.close(fig)
-        
-        return surface
-    
-    def render_price_history(self, history: HistoryData, width: int, height: int) -> pygame.Surface:
-        """Render price history chart for all resources."""
-        fig, ax = plt.subplots(figsize=(width/self.dpi, height/self.dpi), dpi=self.dpi)
-        
-        turns = list(history.turns) if history.turns else [0]
-        
-        has_data = False
-        for i, (resource, prices) in enumerate(history.prices.items()):
-            if prices and any(p is not None for p in prices):
-                color = ChartColors.SERIES[i % len(ChartColors.SERIES)]
-                # Filter out None values
-                valid_prices = [p if p is not None else 0 for p in prices]
-                # Align with turns
-                min_len = min(len(turns), len(valid_prices))
-                ax.plot(list(turns)[-min_len:], valid_prices[-min_len:], 
-                       color=color, linewidth=1.5, label=resource.capitalize(), alpha=0.9)
-                has_data = True
-        
-        ax.set_title('Market Prices', color=ChartColors.CYAN)
-        ax.set_xlabel('Turn')
-        ax.set_ylabel('Price (coins)')
-        ax.grid(True, alpha=0.2)
-        
-        if has_data:
-            ax.legend(loc='upper left', fontsize=8, framealpha=0.8)
-        
-        ax.set_ylim(bottom=0)
-        ax.yaxis.set_major_locator(mticker.MaxNLocator(integer=True))
-        
-        fig.tight_layout()
-        return self._fig_to_surface(fig)
-    
-    def render_population(self, history: HistoryData, width: int, height: int) -> pygame.Surface:
-        """Render population over time chart."""
-        fig, ax = plt.subplots(figsize=(width/self.dpi, height/self.dpi), dpi=self.dpi)
-        
-        turns = list(history.turns) if history.turns else [0]
-        population = list(history.population) if history.population else [0]
-        deaths = list(history.deaths_cumulative) if history.deaths_cumulative else [0]
-        
-        min_len = min(len(turns), len(population))
-        
-        # Population line
-        ax.fill_between(turns[-min_len:], population[-min_len:], 
-                       alpha=0.3, color=ChartColors.CYAN)
-        ax.plot(turns[-min_len:], population[-min_len:], 
-               color=ChartColors.CYAN, linewidth=2, label='Living')
-        
-        # Deaths line
-        if deaths:
-            ax.plot(turns[-min_len:], deaths[-min_len:], 
-                   color=ChartColors.MAGENTA, linewidth=1.5, linestyle='--', 
-                   label='Total Deaths', alpha=0.8)
-        
-        ax.set_title('Population Over Time', color=ChartColors.LIME)
-        ax.set_xlabel('Turn')
-        ax.set_ylabel('Count')
-        ax.grid(True, alpha=0.2)
-        ax.legend(loc='upper right', fontsize=8)
-        ax.set_ylim(bottom=0)
-        ax.yaxis.set_major_locator(mticker.MaxNLocator(integer=True))
-        
-        fig.tight_layout()
-        return self._fig_to_surface(fig)
-    
-    def render_wealth_distribution(self, state: "SimulationState", width: int, height: int) -> pygame.Surface:
-        """Render current wealth distribution as a bar chart."""
-        fig, ax = plt.subplots(figsize=(width/self.dpi, height/self.dpi), dpi=self.dpi)
-        
-        # Get agent wealth data
-        agents = [a for a in state.agents if a.get("is_alive", False)]
-        if not agents:
-            ax.text(0.5, 0.5, 'No living agents', ha='center', va='center', 
-                   color=ChartColors.TEXT_DIM, fontsize=12)
-            ax.set_title('Wealth Distribution', color=ChartColors.ORANGE)
-            fig.tight_layout()
-            return self._fig_to_surface(fig)
-        
-        # Sort by wealth
-        agents_sorted = sorted(agents, key=lambda a: a.get("money", 0), reverse=True)
-        names = [a.get("name", "?")[:8] for a in agents_sorted]
-        wealth = [a.get("money", 0) for a in agents_sorted]
-        
-        # Create gradient colors based on wealth ranking
-        colors = []
-        for i in range(len(agents_sorted)):
-            ratio = i / max(1, len(agents_sorted) - 1)
-            # Gradient from cyan (rich) to magenta (poor)
-            r = int(0 + ratio * 255)
-            g = int(212 - ratio * 212)
-            b = int(255 - ratio * 102)
-            colors.append(f'#{r:02x}{g:02x}{b:02x}')
-        
-        bars = ax.barh(range(len(agents_sorted)), wealth, color=colors, alpha=0.85)
-        ax.set_yticks(range(len(agents_sorted)))
-        ax.set_yticklabels(names, fontsize=7)
-        ax.invert_yaxis()  # Rich at top
-        
-        # Add value labels
-        for bar, val in zip(bars, wealth):
-            ax.text(bar.get_width() + 1, bar.get_y() + bar.get_height()/2,
-                   f'{val}', va='center', fontsize=7, color=ChartColors.TEXT_DIM)
-        
-        ax.set_title('Wealth Distribution', color=ChartColors.ORANGE)
-        ax.set_xlabel('Coins')
-        ax.grid(True, alpha=0.2, axis='x')
-        
-        fig.tight_layout()
-        return self._fig_to_surface(fig)
-    
-    def render_wealth_over_time(self, history: HistoryData, width: int, height: int) -> pygame.Surface:
-        """Render wealth metrics over time (total money, avg, gini)."""
-        fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(width/self.dpi, height/self.dpi), 
-                                       dpi=self.dpi, height_ratios=[2, 1])
-        
-        turns = list(history.turns) if history.turns else [0]
-        total = list(history.total_money) if history.total_money else [0]
-        avg = list(history.avg_wealth) if history.avg_wealth else [0]
-        gini = list(history.gini_coefficient) if history.gini_coefficient else [0]
-        
-        min_len = min(len(turns), len(total), len(avg))
-        
-        # Total and average wealth
-        ax1.plot(turns[-min_len:], total[-min_len:], 
-                color=ChartColors.CYAN, linewidth=2, label='Total Money')
-        ax1.fill_between(turns[-min_len:], total[-min_len:], 
-                        alpha=0.2, color=ChartColors.CYAN)
-        
-        ax1_twin = ax1.twinx()
-        ax1_twin.plot(turns[-min_len:], avg[-min_len:], 
-                     color=ChartColors.LIME, linewidth=1.5, linestyle='--', label='Avg Wealth')
-        ax1_twin.set_ylabel('Avg Wealth', color=ChartColors.LIME)
-        ax1_twin.tick_params(axis='y', labelcolor=ChartColors.LIME)
-        
-        ax1.set_title('Money in Circulation', color=ChartColors.YELLOW)
-        ax1.set_ylabel('Total Money', color=ChartColors.CYAN)
-        ax1.tick_params(axis='y', labelcolor=ChartColors.CYAN)
-        ax1.grid(True, alpha=0.2)
-        ax1.set_ylim(bottom=0)
-        
-        # Gini coefficient (inequality)
-        min_len_gini = min(len(turns), len(gini))
-        ax2.fill_between(turns[-min_len_gini:], gini[-min_len_gini:], 
-                        alpha=0.4, color=ChartColors.MAGENTA)
-        ax2.plot(turns[-min_len_gini:], gini[-min_len_gini:], 
-                color=ChartColors.MAGENTA, linewidth=1.5)
-        ax2.set_xlabel('Turn')
-        ax2.set_ylabel('Gini')
-        ax2.set_title('Inequality Index', color=ChartColors.MAGENTA, fontsize=9)
-        ax2.set_ylim(0, 1)
-        ax2.grid(True, alpha=0.2)
-        
-        # Add reference lines for gini
-        ax2.axhline(y=0.4, color=ChartColors.YELLOW, linestyle=':', alpha=0.5, linewidth=1)
-        ax2.text(turns[-1] if turns else 0, 0.42, 'Moderate', fontsize=7, 
-                color=ChartColors.YELLOW, alpha=0.7)
-        
-        fig.tight_layout()
-        return self._fig_to_surface(fig)
-    
-    def render_professions(self, state: "SimulationState", history: HistoryData, 
-                          width: int, height: int) -> pygame.Surface:
-        """Render profession distribution as pie chart and area chart."""
-        fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(width/self.dpi, height/self.dpi), dpi=self.dpi)
-        
-        # Current profession pie chart
-        professions = state.statistics.get("professions", {})
-        if professions:
-            labels = list(professions.keys())
-            sizes = list(professions.values())
-            colors = [ChartColors.SERIES[i % len(ChartColors.SERIES)] for i in range(len(labels))]
-            
-            wedges, texts, autotexts = ax1.pie(
-                sizes, labels=labels, colors=colors, autopct='%1.0f%%',
-                startangle=90, pctdistance=0.75,
-                textprops={'fontsize': 8, 'color': ChartColors.TEXT}
-            )
-            for autotext in autotexts:
-                autotext.set_color(ChartColors.BG)
-                autotext.set_fontweight('bold')
-            ax1.set_title('Current Distribution', color=ChartColors.PURPLE, fontsize=10)
-        else:
-            ax1.text(0.5, 0.5, 'No data', ha='center', va='center', color=ChartColors.TEXT_DIM)
-            ax1.set_title('Current Distribution', color=ChartColors.PURPLE)
-        
-        # Profession history as stacked area
-        turns = list(history.turns) if history.turns else [0]
-        if history.professions and turns:
-            profs_list = list(history.professions.keys())
-            data = []
-            for prof in profs_list:
-                prof_data = list(history.professions[prof])
-                # Pad to match turns length
-                while len(prof_data) < len(turns):
-                    prof_data.insert(0, 0)
-                data.append(prof_data[-len(turns):])
-            
-            colors = [ChartColors.SERIES[i % len(ChartColors.SERIES)] for i in range(len(profs_list))]
-            ax2.stackplot(turns, *data, labels=profs_list, colors=colors, alpha=0.8)
-            ax2.legend(loc='upper left', fontsize=7, framealpha=0.8)
-            ax2.set_xlabel('Turn')
-            ax2.set_ylabel('Count')
-        
-        ax2.set_title('Over Time', color=ChartColors.PURPLE, fontsize=10)
-        ax2.grid(True, alpha=0.2)
-        
-        fig.tight_layout()
-        return self._fig_to_surface(fig)
-    
-    def render_market_activity(self, state: "SimulationState", history: HistoryData,
-                               width: int, height: int) -> pygame.Surface:
-        """Render market activity - orders by resource, supply/demand."""
-        fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(width/self.dpi, height/self.dpi), dpi=self.dpi)
-        
-        # Current market orders by resource type
-        prices = state.market_prices
-        resources = []
-        quantities = []
-        colors = []
-        
-        for i, (resource, data) in enumerate(prices.items()):
-            qty = data.get("total_available", 0)
-            if qty > 0:
-                resources.append(resource.capitalize())
-                quantities.append(qty)
-                colors.append(ChartColors.SERIES[i % len(ChartColors.SERIES)])
-        
-        if resources:
-            bars = ax1.bar(resources, quantities, color=colors, alpha=0.85)
-            ax1.set_ylabel('Available')
-            for bar, val in zip(bars, quantities):
-                ax1.text(bar.get_x() + bar.get_width()/2, bar.get_height() + 0.3,
-                        str(val), ha='center', fontsize=8, color=ChartColors.TEXT)
-        else:
-            ax1.text(0.5, 0.5, 'No orders', ha='center', va='center', color=ChartColors.TEXT_DIM)
-        
-        ax1.set_title('Market Supply', color=ChartColors.TEAL, fontsize=10)
-        ax1.tick_params(axis='x', rotation=45, labelsize=7)
-        ax1.grid(True, alpha=0.2, axis='y')
-        
-        # Supply/Demand scores
-        resources_sd = []
-        supply_scores = []
-        demand_scores = []
-        
-        for resource, data in prices.items():
-            resources_sd.append(resource[:6])
-            supply_scores.append(data.get("supply_score", 0.5))
-            demand_scores.append(data.get("demand_score", 0.5))
-        
-        if resources_sd:
-            x = np.arange(len(resources_sd))
-            width_bar = 0.35
-            
-            ax2.bar(x - width_bar/2, supply_scores, width_bar, label='Supply', 
-                   color=ChartColors.CYAN, alpha=0.8)
-            ax2.bar(x + width_bar/2, demand_scores, width_bar, label='Demand', 
-                   color=ChartColors.MAGENTA, alpha=0.8)
-            
-            ax2.set_xticks(x)
-            ax2.set_xticklabels(resources_sd, fontsize=7, rotation=45)
-            ax2.set_ylabel('Score')
-            ax2.legend(fontsize=7)
-            ax2.set_ylim(0, 1.2)
-        
-        ax2.set_title('Supply/Demand', color=ChartColors.TEAL, fontsize=10)
-        ax2.grid(True, alpha=0.2, axis='y')
-        
-        fig.tight_layout()
-        return self._fig_to_surface(fig)
-    
-    def render_agent_stats(self, state: "SimulationState", width: int, height: int) -> pygame.Surface:
-        """Render aggregate agent statistics - energy, hunger, thirst distributions."""
-        fig, axes = plt.subplots(2, 2, figsize=(width/self.dpi, height/self.dpi), dpi=self.dpi)
-        
-        agents = [a for a in state.agents if a.get("is_alive", False)]
-        
-        if not agents:
-            for ax in axes.flat:
-                ax.text(0.5, 0.5, 'No agents', ha='center', va='center', color=ChartColors.TEXT_DIM)
-            fig.suptitle('Agent Statistics', color=ChartColors.CYAN)
-            fig.tight_layout()
-            return self._fig_to_surface(fig)
-        
-        # Extract stats
-        energies = [a.get("stats", {}).get("energy", 0) for a in agents]
-        hungers = [a.get("stats", {}).get("hunger", 0) for a in agents]
-        thirsts = [a.get("stats", {}).get("thirst", 0) for a in agents]
-        heats = [a.get("stats", {}).get("heat", 0) for a in agents]
-        
-        max_energy = agents[0].get("stats", {}).get("max_energy", 100)
-        max_hunger = agents[0].get("stats", {}).get("max_hunger", 100)
-        max_thirst = agents[0].get("stats", {}).get("max_thirst", 100)
-        max_heat = agents[0].get("stats", {}).get("max_heat", 100)
-        
-        stats_data = [
-            (energies, max_energy, 'Energy', ChartColors.LIME),
-            (hungers, max_hunger, 'Hunger', ChartColors.ORANGE),
-            (thirsts, max_thirst, 'Thirst', ChartColors.CYAN),
-            (heats, max_heat, 'Heat', ChartColors.MAGENTA),
-        ]
-        
-        for ax, (values, max_val, name, color) in zip(axes.flat, stats_data):
-            # Histogram
-            bins = np.linspace(0, max_val, 11)
-            ax.hist(values, bins=bins, color=color, alpha=0.7, edgecolor=ChartColors.PANEL)
-            
-            # Mean line
-            mean_val = np.mean(values)
-            ax.axvline(x=mean_val, color=ChartColors.TEXT, linestyle='--', 
-                      linewidth=1.5, label=f'Avg: {mean_val:.0f}')
-            
-            # Critical threshold
-            critical = max_val * 0.25
-            ax.axvline(x=critical, color=ChartColors.MAGENTA, linestyle=':', 
-                      linewidth=1, alpha=0.7)
-            
-            ax.set_title(name, color=color, fontsize=9)
-            ax.set_xlim(0, max_val)
-            ax.legend(fontsize=7, loc='upper right')
-            ax.grid(True, alpha=0.2)
-        
-        fig.suptitle('Agent Statistics Distribution', color=ChartColors.CYAN, fontsize=11)
-        fig.tight_layout()
-        return self._fig_to_surface(fig)
-
-
-class StatsRenderer:
-    """Main statistics panel with tabs and charts."""
-    
-    TABS = [
-        ("Prices", "price_history"),
-        ("Wealth", "wealth"),
-        ("Population", "population"),
-        ("Professions", "professions"),
-        ("Market", "market"),
-        ("Agent Stats", "agent_stats"),
-    ]
-    
-    def __init__(self, screen: pygame.Surface):
-        self.screen = screen
-        self.visible = False
-        
-        self.font = pygame.font.Font(None, 24)
-        self.small_font = pygame.font.Font(None, 18)
-        self.title_font = pygame.font.Font(None, 32)
-        
-        self.current_tab = 0
-        self.tab_hovered = -1
-        
-        # History data
-        self.history = HistoryData()
-        
-        # Chart renderer
-        self.chart_renderer: Optional[ChartRenderer] = None
-        
-        # Cached chart surfaces
-        self._chart_cache: dict[str, pygame.Surface] = {}
-        self._cache_turn: int = -1
-        
-        # Layout
-        self._calculate_layout()
-    
-    def _calculate_layout(self) -> None:
-        """Calculate panel layout based on screen size."""
-        screen_w, screen_h = self.screen.get_size()
-        
-        # Panel takes most of the screen with some margin
-        margin = 30
-        self.panel_rect = pygame.Rect(
-            margin, margin,
-            screen_w - margin * 2,
-            screen_h - margin * 2
-        )
-        
-        # Tab bar
-        self.tab_height = 40
-        self.tab_rect = pygame.Rect(
-            self.panel_rect.x,
-            self.panel_rect.y,
-            self.panel_rect.width,
-            self.tab_height
-        )
-        
-        # Chart area
-        self.chart_rect = pygame.Rect(
-            self.panel_rect.x + 10,
-            self.panel_rect.y + self.tab_height + 10,
-            self.panel_rect.width - 20,
-            self.panel_rect.height - self.tab_height - 20
-        )
-        
-        # Initialize chart renderer with chart area size
-        self.chart_renderer = ChartRenderer(
-            self.chart_rect.width,
-            self.chart_rect.height
-        )
-        
-        # Calculate tab widths
-        self.tab_width = self.panel_rect.width // len(self.TABS)
-    
-    def toggle(self) -> None:
-        """Toggle visibility of the stats panel."""
-        self.visible = not self.visible
-        if self.visible:
-            self._invalidate_cache()
-    
-    def update_history(self, state: "SimulationState") -> None:
-        """Update history data with new state."""
-        if state:
-            self.history.update(state)
-    
-    def clear_history(self) -> None:
-        """Clear all history data (e.g., on simulation reset)."""
-        self.history.clear()
-        self._invalidate_cache()
-    
-    def _invalidate_cache(self) -> None:
-        """Invalidate chart cache to force re-render."""
-        self._chart_cache.clear()
-        self._cache_turn = -1
-    
-    def handle_event(self, event: pygame.event.Event) -> bool:
-        """Handle input events. Returns True if event was consumed."""
-        if not self.visible:
-            return False
-        
-        if event.type == pygame.MOUSEMOTION:
-            self._handle_mouse_motion(event.pos)
-            return True
-        
-        elif event.type == pygame.MOUSEBUTTONDOWN:
-            if self._handle_click(event.pos):
-                return True
-            # Consume clicks when visible
-            return True
-        
-        elif event.type == pygame.KEYDOWN:
-            if event.key == pygame.K_ESCAPE:
-                self.toggle()
-                return True
-            elif event.key == pygame.K_LEFT:
-                self.current_tab = (self.current_tab - 1) % len(self.TABS)
-                self._invalidate_cache()
-                return True
-            elif event.key == pygame.K_RIGHT:
-                self.current_tab = (self.current_tab + 1) % len(self.TABS)
-                self._invalidate_cache()
-                return True
-        
-        return False
-    
-    def _handle_mouse_motion(self, pos: tuple[int, int]) -> None:
-        """Handle mouse motion for tab hover effects."""
-        self.tab_hovered = -1
-        
-        if self.tab_rect.collidepoint(pos):
-            rel_x = pos[0] - self.tab_rect.x
-            tab_idx = rel_x // self.tab_width
-            if 0 <= tab_idx < len(self.TABS):
-                self.tab_hovered = tab_idx
-    
-    def _handle_click(self, pos: tuple[int, int]) -> bool:
-        """Handle mouse click. Returns True if click was on a tab."""
-        if self.tab_rect.collidepoint(pos):
-            rel_x = pos[0] - self.tab_rect.x
-            tab_idx = rel_x // self.tab_width
-            if 0 <= tab_idx < len(self.TABS) and tab_idx != self.current_tab:
-                self.current_tab = tab_idx
-                self._invalidate_cache()
-                return True
-        return False
-    
-    def _render_chart(self, state: "SimulationState") -> pygame.Surface:
-        """Render the current tab's chart."""
-        tab_name, tab_key = self.TABS[self.current_tab]
-        
-        # Check cache
-        current_turn = state.turn if state else 0
-        if tab_key in self._chart_cache and self._cache_turn == current_turn:
-            return self._chart_cache[tab_key]
-        
-        # Render chart based on current tab
-        width = self.chart_rect.width
-        height = self.chart_rect.height
-        
-        if tab_key == "price_history":
-            surface = self.chart_renderer.render_price_history(self.history, width, height)
-        elif tab_key == "wealth":
-            # Split into two charts
-            half_height = height // 2
-            dist_surface = self.chart_renderer.render_wealth_distribution(state, width, half_height)
-            time_surface = self.chart_renderer.render_wealth_over_time(self.history, width, half_height)
-            
-            surface = pygame.Surface((width, height))
-            surface.fill(UIColors.BG)
-            surface.blit(dist_surface, (0, 0))
-            surface.blit(time_surface, (0, half_height))
-        elif tab_key == "population":
-            surface = self.chart_renderer.render_population(self.history, width, height)
-        elif tab_key == "professions":
-            surface = self.chart_renderer.render_professions(state, self.history, width, height)
-        elif tab_key == "market":
-            surface = self.chart_renderer.render_market_activity(state, self.history, width, height)
-        elif tab_key == "agent_stats":
-            surface = self.chart_renderer.render_agent_stats(state, width, height)
-        else:
-            # Fallback empty surface
-            surface = pygame.Surface((width, height))
-            surface.fill(UIColors.BG)
-        
-        # Cache the result
-        self._chart_cache[tab_key] = surface
-        self._cache_turn = current_turn
-        
-        return surface
-    
-    def draw(self, state: "SimulationState") -> None:
-        """Draw the statistics panel."""
-        if not self.visible:
-            return
-        
-        # Dim background
-        overlay = pygame.Surface(self.screen.get_size(), pygame.SRCALPHA)
-        overlay.fill((0, 0, 0, 220))
-        self.screen.blit(overlay, (0, 0))
-        
-        # Panel background
-        pygame.draw.rect(self.screen, UIColors.PANEL_BG, self.panel_rect, border_radius=12)
-        pygame.draw.rect(self.screen, UIColors.PANEL_BORDER, self.panel_rect, 2, border_radius=12)
-        
-        # Draw tabs
-        self._draw_tabs()
-        
-        # Draw chart
-        if state:
-            chart_surface = self._render_chart(state)
-            self.screen.blit(chart_surface, self.chart_rect.topleft)
-        
-        # Draw close hint
-        hint = self.small_font.render("Press G or ESC to close  |  ← → to switch tabs", 
-                                      True, UIColors.TEXT_SECONDARY)
-        hint_rect = hint.get_rect(centerx=self.panel_rect.centerx, 
-                                  y=self.panel_rect.bottom - 25)
-        self.screen.blit(hint, hint_rect)
-    
-    def _draw_tabs(self) -> None:
-        """Draw the tab bar."""
-        for i, (tab_name, _) in enumerate(self.TABS):
-            tab_x = self.tab_rect.x + i * self.tab_width
-            tab_rect = pygame.Rect(tab_x, self.tab_rect.y, self.tab_width, self.tab_height)
-            
-            # Tab background
-            if i == self.current_tab:
-                color = UIColors.TAB_ACTIVE
-            elif i == self.tab_hovered:
-                color = UIColors.TAB_HOVER
-            else:
-                color = UIColors.TAB_INACTIVE
-            
-            # Draw tab with rounded top corners
-            tab_surface = pygame.Surface((self.tab_width, self.tab_height), pygame.SRCALPHA)
-            pygame.draw.rect(tab_surface, color, (0, 0, self.tab_width, self.tab_height), 
-                           border_top_left_radius=8, border_top_right_radius=8)
-            
-            if i == self.current_tab:
-                # Active tab - solid color
-                tab_surface.set_alpha(255)
-            else:
-                tab_surface.set_alpha(180)
-            
-            self.screen.blit(tab_surface, (tab_x, self.tab_rect.y))
-            
-            # Tab text
-            text_color = UIColors.BG if i == self.current_tab else UIColors.TEXT_PRIMARY
-            text = self.small_font.render(tab_name, True, text_color)
-            text_rect = text.get_rect(center=tab_rect.center)
-            self.screen.blit(text, text_rect)
-            
-            # Tab border
-            if i != self.current_tab:
-                pygame.draw.line(self.screen, UIColors.PANEL_BORDER,
-                               (tab_x + self.tab_width - 1, self.tab_rect.y + 5),
-                               (tab_x + self.tab_width - 1, self.tab_rect.y + self.tab_height - 5))
-

frontend/renderer/ui_renderer.py

@@ -1,239 +0,0 @@
-"""UI renderer for the Village Simulation."""
-
-import pygame
-from typing import TYPE_CHECKING, Optional
-
-if TYPE_CHECKING:
-    from frontend.client import SimulationState
-
-
-class Colors:
-    # UI colors
-    PANEL_BG = (35, 40, 50)
-    PANEL_BORDER = (70, 80, 95)
-    TEXT_PRIMARY = (230, 230, 235)
-    TEXT_SECONDARY = (160, 165, 175)
-    TEXT_HIGHLIGHT = (100, 180, 255)
-    TEXT_WARNING = (255, 180, 80)
-    TEXT_DANGER = (255, 100, 100)
-    
-    # Day/Night indicator
-    DAY_COLOR = (255, 220, 100)
-    NIGHT_COLOR = (100, 120, 180)
-
-
-class UIRenderer:
-    """Renders UI elements (HUD, panels, text info)."""
-    
-    def __init__(self, screen: pygame.Surface, font: pygame.font.Font):
-        self.screen = screen
-        self.font = font
-        self.small_font = pygame.font.Font(None, 20)
-        self.title_font = pygame.font.Font(None, 28)
-        
-        # Panel dimensions
-        self.top_panel_height = 50
-        self.right_panel_width = 200
-    
-    def _draw_panel(self, rect: pygame.Rect, title: Optional[str] = None) -> None:
-        """Draw a panel background."""
-        pygame.draw.rect(self.screen, Colors.PANEL_BG, rect)
-        pygame.draw.rect(self.screen, Colors.PANEL_BORDER, rect, 1)
-        
-        if title:
-            title_text = self.small_font.render(title, True, Colors.TEXT_SECONDARY)
-            self.screen.blit(title_text, (rect.x + 8, rect.y + 4))
-    
-    def draw_top_bar(self, state: "SimulationState") -> None:
-        """Draw the top information bar."""
-        rect = pygame.Rect(0, 0, self.screen.get_width(), self.top_panel_height)
-        pygame.draw.rect(self.screen, Colors.PANEL_BG, rect)
-        pygame.draw.line(
-            self.screen,
-            Colors.PANEL_BORDER,
-            (0, self.top_panel_height),
-            (self.screen.get_width(), self.top_panel_height),
-        )
-        
-        # Day/Night and Turn info
-        is_night = state.time_of_day == "night"
-        time_color = Colors.NIGHT_COLOR if is_night else Colors.DAY_COLOR
-        time_text = "NIGHT" if is_night else "DAY"
-        
-        # Draw time indicator circle
-        pygame.draw.circle(self.screen, time_color, (25, 25), 12)
-        pygame.draw.circle(self.screen, Colors.PANEL_BORDER, (25, 25), 12, 1)
-        
-        # Time/day text
-        info_text = f"{time_text}  |  Day {state.day}, Step {state.step_in_day}  |  Turn {state.turn}"
-        text = self.font.render(info_text, True, Colors.TEXT_PRIMARY)
-        self.screen.blit(text, (50, 15))
-        
-        # Mode indicator
-        mode_color = Colors.TEXT_HIGHLIGHT if state.mode == "auto" else Colors.TEXT_SECONDARY
-        mode_text = f"Mode: {state.mode.upper()}"
-        text = self.small_font.render(mode_text, True, mode_color)
-        self.screen.blit(text, (self.screen.get_width() - 120, 8))
-        
-        # Running indicator
-        if state.is_running:
-            status_text = "RUNNING"
-            status_color = (100, 200, 100)
-        else:
-            status_text = "STOPPED"
-            status_color = Colors.TEXT_DANGER
-        
-        text = self.small_font.render(status_text, True, status_color)
-        self.screen.blit(text, (self.screen.get_width() - 120, 28))
-    
-    def draw_right_panel(self, state: "SimulationState") -> None:
-        """Draw the right information panel."""
-        panel_x = self.screen.get_width() - self.right_panel_width
-        rect = pygame.Rect(
-            panel_x,
-            self.top_panel_height,
-            self.right_panel_width,
-            self.screen.get_height() - self.top_panel_height,
-        )
-        pygame.draw.rect(self.screen, Colors.PANEL_BG, rect)
-        pygame.draw.line(
-            self.screen,
-            Colors.PANEL_BORDER,
-            (panel_x, self.top_panel_height),
-            (panel_x, self.screen.get_height()),
-        )
-        
-        y = self.top_panel_height + 10
-        
-        # Statistics section
-        y = self._draw_statistics_section(state, panel_x + 10, y)
-        
-        # Market section
-        y = self._draw_market_section(state, panel_x + 10, y + 20)
-        
-        # Controls help section
-        self._draw_controls_help(panel_x + 10, self.screen.get_height() - 100)
-    
-    def _draw_statistics_section(self, state: "SimulationState", x: int, y: int) -> int:
-        """Draw the statistics section."""
-        # Title
-        title = self.title_font.render("Statistics", True, Colors.TEXT_PRIMARY)
-        self.screen.blit(title, (x, y))
-        y += 30
-        
-        stats = state.statistics
-        living = len(state.get_living_agents())
-        
-        # Population
-        pop_color = Colors.TEXT_PRIMARY if living > 2 else Colors.TEXT_DANGER
-        text = self.small_font.render(f"Population: {living}", True, pop_color)
-        self.screen.blit(text, (x, y))
-        y += 18
-        
-        # Deaths
-        deaths = stats.get("total_agents_died", 0)
-        if deaths > 0:
-            text = self.small_font.render(f"Deaths: {deaths}", True, Colors.TEXT_WARNING)
-            self.screen.blit(text, (x, y))
-            y += 18
-        
-        # Total money
-        total_money = stats.get("total_money_in_circulation", 0)
-        text = self.small_font.render(f"Total Coins: {total_money}", True, Colors.TEXT_SECONDARY)
-        self.screen.blit(text, (x, y))
-        y += 18
-        
-        # Professions
-        professions = stats.get("professions", {})
-        if professions:
-            y += 5
-            text = self.small_font.render("Professions:", True, Colors.TEXT_SECONDARY)
-            self.screen.blit(text, (x, y))
-            y += 16
-            
-            for prof, count in professions.items():
-                text = self.small_font.render(f"  {prof}: {count}", True, Colors.TEXT_SECONDARY)
-                self.screen.blit(text, (x, y))
-                y += 14
-        
-        return y
-    
-    def _draw_market_section(self, state: "SimulationState", x: int, y: int) -> int:
-        """Draw the market section."""
-        # Title
-        title = self.title_font.render("Market", True, Colors.TEXT_PRIMARY)
-        self.screen.blit(title, (x, y))
-        y += 30
-        
-        # Order count
-        order_count = len(state.market_orders)
-        text = self.small_font.render(f"Active Orders: {order_count}", True, Colors.TEXT_SECONDARY)
-        self.screen.blit(text, (x, y))
-        y += 20
-        
-        # Price summary for each resource with available stock
-        prices = state.market_prices
-        for resource, data in prices.items():
-            if data.get("total_available", 0) > 0:
-                price = data.get("lowest_price", "?")
-                qty = data.get("total_available", 0)
-                text = self.small_font.render(
-                    f"{resource}: {qty}x @ {price}c",
-                    True,
-                    Colors.TEXT_SECONDARY,
-                )
-                self.screen.blit(text, (x, y))
-                y += 16
-        
-        return y
-    
-    def _draw_controls_help(self, x: int, y: int) -> None:
-        """Draw controls help at bottom of panel."""
-        pygame.draw.line(
-            self.screen,
-            Colors.PANEL_BORDER,
-            (x - 5, y - 10),
-            (self.screen.get_width() - 5, y - 10),
-        )
-        
-        title = self.small_font.render("Controls", True, Colors.TEXT_PRIMARY)
-        self.screen.blit(title, (x, y))
-        y += 20
-        
-        controls = [
-            "SPACE - Next Turn",
-            "R - Reset Simulation",
-            "M - Toggle Mode",
-            "S - Settings",
-            "ESC - Quit",
-        ]
-        
-        for control in controls:
-            text = self.small_font.render(control, True, Colors.TEXT_SECONDARY)
-            self.screen.blit(text, (x, y))
-            y += 16
-    
-    def draw_connection_status(self, connected: bool) -> None:
-        """Draw connection status overlay when disconnected."""
-        if connected:
-            return
-        
-        # Semi-transparent overlay
-        overlay = pygame.Surface(self.screen.get_size(), pygame.SRCALPHA)
-        overlay.fill((0, 0, 0, 180))
-        self.screen.blit(overlay, (0, 0))
-        
-        # Connection message
-        text = self.title_font.render("Connecting to server...", True, Colors.TEXT_WARNING)
-        text_rect = text.get_rect(center=(self.screen.get_width() // 2, self.screen.get_height() // 2))
-        self.screen.blit(text, text_rect)
-        
-        hint = self.small_font.render("Make sure the backend is running on localhost:8000", True, Colors.TEXT_SECONDARY)
-        hint_rect = hint.get_rect(center=(self.screen.get_width() // 2, self.screen.get_height() // 2 + 30))
-        self.screen.blit(hint, hint_rect)
-    
-    def draw(self, state: "SimulationState") -> None:
-        """Draw all UI elements."""
-        self.draw_top_bar(state)
-        self.draw_right_panel(state)
-

requirements.txt

@@ -5,8 +5,7 @@ fastapi>=0.104.0
 uvicorn[standard]>=0.24.0
 pydantic>=2.5.0
 
-# Frontend
-pygame-ce>=2.4.0
+# HTTP client (for web frontend communication)
 requests>=2.31.0
 
 # Tools (balance sheet export/import)

tools/optimize_goap.py

@@ -0,0 +1,496 @@
+#!/usr/bin/env python3
+"""
+GOAP Economy Optimizer for Village Simulation
+
+This script optimizes the simulation parameters specifically for the GOAP AI system.
+The goal is to achieve:
+- Balanced action diversity (hunting, gathering, trading)
+- Active economy with trading
+- Good survival rates
+- Meat production through hunting
+
+Key insight: GOAP uses action COSTS to choose actions. Lower cost = preferred.
+We need to tune:
+1. Action energy costs (config.json)
+2. GOAP action cost functions (goap/actions.py)
+3. Goal priorities (goap/goals.py)
+
+Usage:
+    python tools/optimize_goap.py [--iterations 15] [--steps 300]
+    python tools/optimize_goap.py --analyze  # Analyze current GOAP behavior
+"""
+
+import argparse
+import json
+import random
+import sys
+from collections import defaultdict
+from datetime import datetime
+from pathlib import Path
+
+# Add parent directory for imports
+sys.path.insert(0, str(Path(__file__).parent.parent))
+
+from backend.config import get_config, reload_config
+from backend.core.engine import GameEngine
+from backend.domain.action import reset_action_config_cache
+from backend.domain.resources import reset_resource_cache
+
+
+def analyze_goap_behavior(num_steps: int = 100, num_agents: int = 10):
+    """Analyze current GOAP behavior in detail."""
+    print("\n" + "=" * 70)
+    print("🔍 GOAP BEHAVIOR ANALYSIS")
+    print("=" * 70)
+
+    # Reset engine
+    GameEngine._instance = None
+    engine = GameEngine()
+    engine.initialize(num_agents=num_agents)
+
+    # Track statistics
+    action_counts = defaultdict(int)
+    goal_counts = defaultdict(int)
+    reactive_count = 0
+    planned_count = 0
+
+    # Resource tracking
+    resources_produced = defaultdict(int)
+    resources_consumed = defaultdict(int)
+
+    # Run simulation
+    for step in range(num_steps):
+        if not engine.is_running:
+            print(f"  Simulation ended at step {step}")
+            break
+
+        log = engine.next_step()
+
+        for action_data in log.agent_actions:
+            decision = action_data.get("decision", {})
+            result = action_data.get("result", {})
+
+            action_type = decision.get("action", "unknown")
+            action_counts[action_type] += 1
+
+            # Track goal/reactive
+            goal_name = decision.get("goal_name", "")
+            reason = decision.get("reason", "")
+
+            if goal_name:
+                goal_counts[goal_name] += 1
+                planned_count += 1
+            elif "Reactive" in reason:
+                goal_counts["(reactive)"] += 1
+                reactive_count += 1
+
+            # Track resources
+            if result and result.get("success"):
+                for res in result.get("resources_gained", []):
+                    resources_produced[res.get("type", "unknown")] += res.get("quantity", 0)
+                for res in result.get("resources_consumed", []):
+                    resources_consumed[res.get("type", "unknown")] += res.get("quantity", 0)
+
+    # Print results
+    total_actions = sum(action_counts.values())
+
+    print(f"\n📊 Action Distribution ({num_steps} turns, {num_agents} agents)")
+    print("-" * 50)
+    for action, count in sorted(action_counts.items(), key=lambda x: -x[1]):
+        pct = count * 100 / total_actions if total_actions > 0 else 0
+        bar = "█" * int(pct / 2)
+        print(f"  {action:12} {count:4} ({pct:5.1f}%) {bar}")
+
+    print(f"\n🎯 Goal Distribution")
+    print("-" * 50)
+    total_goals = sum(goal_counts.values())
+    for goal, count in sorted(goal_counts.items(), key=lambda x: -x[1])[:15]:
+        pct = count * 100 / total_goals if total_goals > 0 else 0
+        print(f"  {goal:20} {count:4} ({pct:5.1f}%)")
+
+    print(f"\n  Planned actions: {planned_count} ({planned_count*100/total_actions:.1f}%)")
+    print(f"  Reactive actions: {reactive_count} ({reactive_count*100/total_actions:.1f}%)")
+
+    print(f"\n📦 Resources Produced")
+    print("-" * 50)
+    for res, qty in sorted(resources_produced.items(), key=lambda x: -x[1]):
+        print(f"  {res:12} {qty:4}")
+
+    print(f"\n🔥 Resources Consumed")
+    print("-" * 50)
+    for res, qty in sorted(resources_consumed.items(), key=lambda x: -x[1]):
+        print(f"  {res:12} {qty:4}")
+
+    # Diagnose issues
+    print(f"\n⚠️  ISSUES DETECTED:")
+    print("-" * 50)
+
+    hunt_pct = action_counts.get("hunt", 0) * 100 / total_actions if total_actions > 0 else 0
+    gather_pct = action_counts.get("gather", 0) * 100 / total_actions if total_actions > 0 else 0
+
+    if hunt_pct < 5:
+        print("  ❌ Almost no hunting! Hunt action cost too high or meat not valued enough.")
+        print("     → Reduce hunt energy cost or increase meat benefits")
+
+    if resources_produced.get("meat", 0) == 0:
+        print("  ❌ No meat produced! Agents never hunt successfully.")
+
+    trade_pct = action_counts.get("trade", 0) * 100 / total_actions if total_actions > 0 else 0
+    if trade_pct < 5:
+        print("  ❌ Low trading activity. Market goals not prioritized.")
+
+    if reactive_count > planned_count:
+        print("  ⚠️  More reactive than planned actions. Goals may be too easily satisfied.")
+
+    return {
+        "action_counts": dict(action_counts),
+        "goal_counts": dict(goal_counts),
+        "resources_produced": dict(resources_produced),
+        "resources_consumed": dict(resources_consumed),
+    }
+
+
+def test_config(config_overrides: dict, num_steps: int = 200, num_agents: int = 10, verbose: bool = True):
+    """Test a configuration and return metrics."""
+    # Save original config
+    config_path = Path("config.json")
+    with open(config_path) as f:
+        original_config = json.load(f)
+
+    # Apply overrides
+    test_config = json.loads(json.dumps(original_config))
+    for section, values in config_overrides.items():
+        if section in test_config:
+            test_config[section].update(values)
+        else:
+            test_config[section] = values
+
+    # Save temp config
+    temp_path = Path("config_temp.json")
+    with open(temp_path, 'w') as f:
+        json.dump(test_config, f, indent=2)
+
+    # Reload config
+    reload_config(str(temp_path))
+    reset_action_config_cache()
+    reset_resource_cache()
+
+    # Run simulation
+    GameEngine._instance = None
+    engine = GameEngine()
+    engine.initialize(num_agents=num_agents)
+
+    action_counts = defaultdict(int)
+    resources_produced = defaultdict(int)
+    deaths = 0
+    trades_completed = 0
+
+    for step in range(num_steps):
+        if not engine.is_running:
+            break
+
+        log = engine.next_step()
+        deaths += len(log.deaths)
+
+        for action_data in log.agent_actions:
+            decision = action_data.get("decision", {})
+            result = action_data.get("result", {})
+
+            action_type = decision.get("action", "unknown")
+            action_counts[action_type] += 1
+
+            if result and result.get("success"):
+                for res in result.get("resources_gained", []):
+                    resources_produced[res.get("type", "unknown")] += res.get("quantity", 0)
+
+                if action_type == "trade" and "Bought" in result.get("message", ""):
+                    trades_completed += 1
+
+    final_pop = len(engine.world.get_living_agents())
+
+    # Cleanup
+    engine.logger.close()
+    temp_path.unlink(missing_ok=True)
+
+    # Restore original config
+    reload_config(str(config_path))
+    reset_action_config_cache()
+    reset_resource_cache()
+
+    # Calculate score
+    total_actions = sum(action_counts.values())
+    hunt_ratio = action_counts.get("hunt", 0) / total_actions if total_actions > 0 else 0
+    gather_ratio = action_counts.get("gather", 0) / total_actions if total_actions > 0 else 0
+    trade_ratio = action_counts.get("trade", 0) / total_actions if total_actions > 0 else 0
+
+    survival_rate = final_pop / num_agents
+
+    # Score components
+    # 1. Hunt ratio: want 10-25%
+    hunt_score = min(25, hunt_ratio * 100) if hunt_ratio > 0.05 else 0
+
+    # 2. Trade activity: want 5-15%
+    trade_score = min(20, trade_ratio * 100 * 2)
+
+    # 3. Resource diversity
+    has_meat = resources_produced.get("meat", 0) > 0
+    has_berries = resources_produced.get("berries", 0) > 0
+    has_wood = resources_produced.get("wood", 0) > 0
+    has_water = resources_produced.get("water", 0) > 0
+    diversity_score = (int(has_meat) + int(has_berries) + int(has_wood) + int(has_water)) * 5
+
+    # 4. Survival
+    survival_score = survival_rate * 30
+
+    # 5. Meat production bonus
+    meat_score = min(15, resources_produced.get("meat", 0) / 5)
+
+    total_score = hunt_score + trade_score + diversity_score + survival_score + meat_score
+
+    if verbose:
+        print(f"\n  Score: {total_score:.1f}/100")
+        print(f"  ├─ Hunt: {hunt_ratio*100:.1f}% ({hunt_score:.1f} pts)")
+        print(f"  ├─ Trade: {trade_ratio*100:.1f}% ({trade_score:.1f} pts)")
+        print(f"  ├─ Diversity: {diversity_score:.1f} pts")
+        print(f"  ├─ Survival: {survival_rate*100:.0f}% ({survival_score:.1f} pts)")
+        print(f"  └─ Meat produced: {resources_produced.get('meat', 0)} ({meat_score:.1f} pts)")
+        print(f"  Actions: hunt={action_counts.get('hunt',0)}, gather={action_counts.get('gather',0)}, trade={action_counts.get('trade',0)}")
+
+    return {
+        "score": total_score,
+        "action_counts": dict(action_counts),
+        "resources": dict(resources_produced),
+        "survival_rate": survival_rate,
+        "deaths": deaths,
+    }
+
+
+def optimize_for_goap(iterations: int = 15, steps: int = 300):
+    """Run optimization focused on GOAP-specific parameters."""
+    print("\n" + "=" * 70)
+    print("🧬 GOAP ECONOMY OPTIMIZER")
+    print("=" * 70)
+    print(f"  Iterations: {iterations}")
+    print(f"  Steps per test: {steps}")
+    print("=" * 70)
+
+    # Key parameters to optimize for GOAP
+    # Focus on making hunting more attractive
+
+    configs_to_test = [
+        # Baseline
+        {
+            "name": "Baseline (current)",
+            "config": {}
+        },
+        # Cheaper hunting
+        {
+            "name": "Cheaper Hunt (-5 energy)",
+            "config": {
+                "actions": {
+                    "hunt_energy": -5,
+                    "hunt_success": 0.8,
+                }
+            }
+        },
+        # More valuable meat
+        {
+            "name": "Valuable Meat (+45 hunger)",
+            "config": {
+                "resources": {
+                    "meat_hunger": 45,
+                    "meat_energy": 15,
+                },
+                "actions": {
+                    "hunt_energy": -6,
+                    "hunt_success": 0.8,
+                }
+            }
+        },
+        # Make berries less attractive
+        {
+            "name": "Nerfed Berries",
+            "config": {
+                "resources": {
+                    "meat_hunger": 45,
+                    "meat_energy": 15,
+                    "berries_hunger": 8,
+                    "berries_thirst": 2,
+                },
+                "actions": {
+                    "hunt_energy": -5,
+                    "gather_energy": -4,
+                    "hunt_success": 0.85,
+                    "hunt_meat_min": 2,
+                    "hunt_meat_max": 4,
+                }
+            }
+        },
+        # Higher hunt output
+        {
+            "name": "High Hunt Output",
+            "config": {
+                "resources": {
+                    "meat_hunger": 40,
+                    "meat_energy": 12,
+                },
+                "actions": {
+                    "hunt_energy": -6,
+                    "hunt_success": 0.85,
+                    "hunt_meat_min": 3,
+                    "hunt_meat_max": 6,
+                    "hunt_hide_min": 1,
+                    "hunt_hide_max": 2,
+                }
+            }
+        },
+        # Balanced economy
+        {
+            "name": "Balanced Economy",
+            "config": {
+                "resources": {
+                    "meat_hunger": 40,
+                    "meat_energy": 15,
+                    "berries_hunger": 8,
+                },
+                "actions": {
+                    "hunt_energy": -5,
+                    "gather_energy": -4,
+                    "hunt_success": 0.8,
+                    "hunt_meat_min": 2,
+                    "hunt_meat_max": 5,
+                },
+                "economy": {
+                    "buy_efficiency_threshold": 0.9,
+                    "min_wealth_target": 40,
+                }
+            }
+        },
+        # Pro-hunting config
+        {
+            "name": "Pro-Hunting",
+            "config": {
+                "agent_stats": {
+                    "hunger_decay": 3,  # Higher hunger decay = need more food
+                },
+                "resources": {
+                    "meat_hunger": 50,  # Meat is very filling
+                    "meat_energy": 15,
+                    "berries_hunger": 6,  # Berries less filling
+                },
+                "actions": {
+                    "hunt_energy": -4,  # Very cheap to hunt
+                    "gather_energy": -4,
+                    "hunt_success": 0.85,
+                    "hunt_meat_min": 3,
+                    "hunt_meat_max": 5,
+                }
+            }
+        },
+        # Full rebalance
+        {
+            "name": "Full Rebalance",
+            "config": {
+                "agent_stats": {
+                    "start_hunger": 70,
+                    "hunger_decay": 3,
+                    "thirst_decay": 3,
+                },
+                "resources": {
+                    "meat_hunger": 50,
+                    "meat_energy": 15,
+                    "berries_hunger": 8,
+                    "berries_thirst": 3,
+                    "water_thirst": 45,
+                },
+                "actions": {
+                    "hunt_energy": -5,
+                    "gather_energy": -4,
+                    "chop_wood_energy": -5,
+                    "get_water_energy": -3,
+                    "hunt_success": 0.8,
+                    "hunt_meat_min": 2,
+                    "hunt_meat_max": 5,
+                    "hunt_hide_min": 0,
+                    "hunt_hide_max": 1,
+                    "gather_min": 2,
+                    "gather_max": 3,
+                }
+            }
+        },
+    ]
+
+    best_config = None
+    best_score = 0
+    best_name = ""
+
+    for cfg in configs_to_test:
+        print(f"\n🧪 Testing: {cfg['name']}")
+        print("-" * 50)
+
+        result = test_config(cfg["config"], steps, verbose=True)
+
+        if result["score"] > best_score:
+            best_score = result["score"]
+            best_config = cfg["config"]
+            best_name = cfg["name"]
+            print(f"  ⭐ New best!")
+
+    print("\n" + "=" * 70)
+    print("🏆 OPTIMIZATION COMPLETE")
+    print("=" * 70)
+    print(f"\n  Best Config: {best_name}")
+    print(f"  Best Score: {best_score:.1f}/100")
+
+    if best_config:
+        print("\n  📝 Configuration to apply:")
+        print("-" * 50)
+        print(json.dumps(best_config, indent=2))
+
+        # Ask to apply
+        print("\n  Would you like to apply this configuration? (y/n)")
+
+        # Save as optimized config
+        output_path = Path("config_goap_optimized.json")
+        with open("config.json") as f:
+            full_config = json.load(f)
+
+        for section, values in best_config.items():
+            if section in full_config:
+                full_config[section].update(values)
+            else:
+                full_config[section] = values
+
+        with open(output_path, 'w') as f:
+            json.dump(full_config, f, indent=2)
+
+        print(f"\n  ✅ Saved to: {output_path}")
+        print("  To apply: cp config_goap_optimized.json config.json")
+
+    return best_config
+
+
+def main():
+    parser = argparse.ArgumentParser(description="Optimize GOAP economy parameters")
+    parser.add_argument("--analyze", "-a", action="store_true", help="Analyze current behavior")
+    parser.add_argument("--iterations", "-i", type=int, default=15, help="Optimization iterations")
+    parser.add_argument("--steps", "-s", type=int, default=200, help="Steps per simulation")
+    parser.add_argument("--apply", action="store_true", help="Auto-apply best config")
+
+    args = parser.parse_args()
+
+    if args.analyze:
+        analyze_goap_behavior(args.steps)
+    else:
+        best = optimize_for_goap(args.iterations, args.steps)
+
+        if args.apply and best:
+            # Apply the config
+            import shutil
+            shutil.copy("config_goap_optimized.json", "config.json")
+            print("\n  ✅ Configuration applied!")
+
+
+if __name__ == "__main__":
+    main()
+

web_frontend/goap_debug.html

@@ -0,0 +1,820 @@
+<!DOCTYPE html>
+<html lang="en">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>GOAP Debug Visualizer - VillSim</title>
+    <link rel="preconnect" href="https://fonts.googleapis.com">
+    <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
+    <link href="https://fonts.googleapis.com/css2?family=IBM+Plex+Sans:wght@400;500;600&family=IBM+Plex+Mono:wght@400;500&display=swap" rel="stylesheet">
+    <script src="https://cdn.jsdelivr.net/npm/chart.js@4.4.1/dist/chart.umd.min.js"></script>
+    <style>
+        :root {
+            --bg-primary: #0d1117;
+            --bg-secondary: #161b22;
+            --bg-tertiary: #21262d;
+            --border-color: #30363d;
+            --text-primary: #e6edf3;
+            --text-secondary: #8b949e;
+            --text-muted: #6e7681;
+            --accent-blue: #58a6ff;
+            --accent-green: #3fb950;
+            --accent-orange: #d29922;
+            --accent-red: #f85149;
+            --accent-purple: #a371f7;
+            --accent-cyan: #39c5cf;
+        }
+
+        * {
+            margin: 0;
+            padding: 0;
+            box-sizing: border-box;
+        }
+
+        body {
+            font-family: 'IBM Plex Sans', -apple-system, BlinkMacSystemFont, sans-serif;
+            background: var(--bg-primary);
+            color: var(--text-primary);
+            min-height: 100vh;
+        }
+
+        .header {
+            background: var(--bg-secondary);
+            border-bottom: 1px solid var(--border-color);
+            padding: 16px 24px;
+            display: flex;
+            align-items: center;
+            justify-content: space-between;
+        }
+
+        .header h1 {
+            font-size: 20px;
+            font-weight: 600;
+            color: var(--accent-cyan);
+        }
+
+        .header-controls {
+            display: flex;
+            gap: 12px;
+            align-items: center;
+        }
+
+        .btn {
+            padding: 8px 16px;
+            border: 1px solid var(--border-color);
+            border-radius: 6px;
+            background: var(--bg-tertiary);
+            color: var(--text-primary);
+            font-family: inherit;
+            font-size: 14px;
+            cursor: pointer;
+            transition: all 0.15s ease;
+        }
+
+        .btn:hover {
+            background: var(--border-color);
+            border-color: var(--text-muted);
+        }
+
+        .btn-primary {
+            background: var(--accent-blue);
+            border-color: var(--accent-blue);
+            color: #fff;
+        }
+
+        .btn-primary:hover {
+            background: #4c9aff;
+        }
+
+        .status-badge {
+            padding: 4px 12px;
+            border-radius: 20px;
+            font-size: 12px;
+            font-weight: 500;
+        }
+
+        .status-connected {
+            background: rgba(63, 185, 80, 0.2);
+            color: var(--accent-green);
+        }
+
+        .status-disconnected {
+            background: rgba(248, 81, 73, 0.2);
+            color: var(--accent-red);
+        }
+
+        .main-content {
+            display: grid;
+            grid-template-columns: 280px 1fr 400px;
+            height: calc(100vh - 65px);
+        }
+
+        .panel {
+            background: var(--bg-secondary);
+            border-right: 1px solid var(--border-color);
+            overflow-y: auto;
+        }
+
+        .panel:last-child {
+            border-right: none;
+            border-left: 1px solid var(--border-color);
+        }
+
+        .panel-header {
+            padding: 16px;
+            border-bottom: 1px solid var(--border-color);
+            background: var(--bg-tertiary);
+            position: sticky;
+            top: 0;
+            z-index: 10;
+        }
+
+        .panel-header h2 {
+            font-size: 14px;
+            font-weight: 600;
+            text-transform: uppercase;
+            letter-spacing: 0.5px;
+            color: var(--text-secondary);
+        }
+
+        .agent-list {
+            padding: 8px;
+        }
+
+        .agent-item {
+            padding: 12px;
+            border-radius: 8px;
+            cursor: pointer;
+            margin-bottom: 4px;
+            transition: background 0.15s ease;
+        }
+
+        .agent-item:hover {
+            background: var(--bg-tertiary);
+        }
+
+        .agent-item.selected {
+            background: rgba(88, 166, 255, 0.15);
+            border: 1px solid var(--accent-blue);
+        }
+
+        .agent-item .agent-name {
+            font-weight: 500;
+            margin-bottom: 4px;
+        }
+
+        .agent-item .agent-action {
+            font-size: 12px;
+            color: var(--text-secondary);
+            font-family: 'IBM Plex Mono', monospace;
+        }
+
+        .agent-item .agent-goal {
+            font-size: 11px;
+            color: var(--accent-cyan);
+            margin-top: 4px;
+        }
+
+        .center-panel {
+            display: flex;
+            flex-direction: column;
+            background: var(--bg-primary);
+        }
+
+        .plan-view {
+            padding: 24px;
+            flex: 1;
+            overflow-y: auto;
+        }
+
+        .plan-header {
+            display: flex;
+            align-items: center;
+            gap: 16px;
+            margin-bottom: 24px;
+        }
+
+        .plan-header h2 {
+            font-size: 24px;
+            font-weight: 600;
+        }
+
+        .plan-goal-badge {
+            padding: 6px 16px;
+            background: rgba(163, 113, 247, 0.2);
+            color: var(--accent-purple);
+            border-radius: 20px;
+            font-size: 14px;
+            font-weight: 500;
+        }
+
+        .world-state-grid {
+            display: grid;
+            grid-template-columns: repeat(4, 1fr);
+            gap: 12px;
+            margin-bottom: 32px;
+        }
+
+        .state-card {
+            background: var(--bg-secondary);
+            border: 1px solid var(--border-color);
+            border-radius: 8px;
+            padding: 16px;
+        }
+
+        .state-card .label {
+            font-size: 11px;
+            color: var(--text-muted);
+            text-transform: uppercase;
+            letter-spacing: 0.5px;
+            margin-bottom: 8px;
+        }
+
+        .state-card .value {
+            font-size: 24px;
+            font-weight: 600;
+            font-family: 'IBM Plex Mono', monospace;
+        }
+
+        .state-card .bar {
+            height: 4px;
+            background: var(--bg-tertiary);
+            border-radius: 2px;
+            margin-top: 8px;
+            overflow: hidden;
+        }
+
+        .state-card .bar-fill {
+            height: 100%;
+            border-radius: 2px;
+            transition: width 0.3s ease;
+        }
+
+        .bar-thirst .bar-fill { background: var(--accent-blue); }
+        .bar-hunger .bar-fill { background: var(--accent-orange); }
+        .bar-heat .bar-fill { background: var(--accent-red); }
+        .bar-energy .bar-fill { background: var(--accent-green); }
+
+        .plan-visualization {
+            background: var(--bg-secondary);
+            border: 1px solid var(--border-color);
+            border-radius: 12px;
+            padding: 24px;
+            margin-bottom: 24px;
+        }
+
+        .plan-visualization h3 {
+            font-size: 14px;
+            color: var(--text-secondary);
+            margin-bottom: 16px;
+        }
+
+        .plan-steps {
+            display: flex;
+            align-items: center;
+            gap: 8px;
+            flex-wrap: wrap;
+        }
+
+        .plan-step {
+            display: flex;
+            align-items: center;
+            gap: 8px;
+        }
+
+        .step-node {
+            padding: 12px 20px;
+            background: var(--bg-tertiary);
+            border: 2px solid var(--border-color);
+            border-radius: 8px;
+            font-family: 'IBM Plex Mono', monospace;
+            font-size: 14px;
+            font-weight: 500;
+        }
+
+        .step-node.current {
+            border-color: var(--accent-green);
+            background: rgba(63, 185, 80, 0.15);
+            color: var(--accent-green);
+        }
+
+        .step-arrow {
+            color: var(--text-muted);
+            font-size: 20px;
+        }
+
+        .goal-result {
+            padding: 12px 20px;
+            background: rgba(163, 113, 247, 0.15);
+            border: 2px solid var(--accent-purple);
+            border-radius: 8px;
+            color: var(--accent-purple);
+            font-weight: 500;
+        }
+
+        .no-plan {
+            text-align: center;
+            padding: 40px;
+            color: var(--text-muted);
+        }
+
+        .goals-chart-container {
+            background: var(--bg-secondary);
+            border: 1px solid var(--border-color);
+            border-radius: 12px;
+            padding: 24px;
+        }
+
+        .goals-chart-container h3 {
+            font-size: 14px;
+            color: var(--text-secondary);
+            margin-bottom: 16px;
+        }
+
+        .chart-wrapper {
+            height: 300px;
+        }
+
+        .detail-section {
+            padding: 16px;
+            border-bottom: 1px solid var(--border-color);
+        }
+
+        .detail-section h3 {
+            font-size: 12px;
+            color: var(--text-muted);
+            text-transform: uppercase;
+            letter-spacing: 0.5px;
+            margin-bottom: 12px;
+        }
+
+        .detail-grid {
+            display: grid;
+            grid-template-columns: 1fr 1fr;
+            gap: 8px;
+        }
+
+        .detail-item {
+            display: flex;
+            justify-content: space-between;
+            padding: 6px 0;
+            font-size: 13px;
+        }
+
+        .detail-item .label {
+            color: var(--text-secondary);
+        }
+
+        .detail-item .value {
+            font-family: 'IBM Plex Mono', monospace;
+            color: var(--text-primary);
+        }
+
+        .action-list {
+            max-height: 300px;
+            overflow-y: auto;
+        }
+
+        .action-item {
+            display: flex;
+            align-items: center;
+            padding: 8px 12px;
+            border-radius: 6px;
+            margin-bottom: 4px;
+            font-size: 13px;
+        }
+
+        .action-item.valid {
+            background: var(--bg-tertiary);
+        }
+
+        .action-item.invalid {
+            background: transparent;
+            opacity: 0.5;
+        }
+
+        .action-item.in-plan {
+            background: rgba(63, 185, 80, 0.15);
+            border: 1px solid var(--accent-green);
+        }
+
+        .action-item .action-name {
+            flex: 1;
+            font-family: 'IBM Plex Mono', monospace;
+        }
+
+        .action-item .action-cost {
+            font-size: 11px;
+            color: var(--text-muted);
+            margin-left: 8px;
+        }
+
+        .action-item .action-order {
+            width: 20px;
+            height: 20px;
+            background: var(--accent-green);
+            border-radius: 50%;
+            display: flex;
+            align-items: center;
+            justify-content: center;
+            font-size: 11px;
+            font-weight: 600;
+            color: #000;
+            margin-right: 8px;
+        }
+
+        .inventory-grid {
+            display: grid;
+            grid-template-columns: repeat(2, 1fr);
+            gap: 8px;
+        }
+
+        .inv-item {
+            display: flex;
+            align-items: center;
+            gap: 8px;
+            padding: 8px;
+            background: var(--bg-tertiary);
+            border-radius: 6px;
+            font-size: 13px;
+        }
+
+        .inv-item .icon {
+            font-size: 16px;
+        }
+
+        .inv-item .count {
+            margin-left: auto;
+            font-family: 'IBM Plex Mono', monospace;
+            font-weight: 500;
+        }
+
+        .loading {
+            display: flex;
+            align-items: center;
+            justify-content: center;
+            height: 100%;
+            color: var(--text-muted);
+        }
+
+        .urgency-indicator {
+            display: inline-block;
+            width: 8px;
+            height: 8px;
+            border-radius: 50%;
+            margin-left: 8px;
+        }
+
+        .urgency-none { background: var(--accent-green); }
+        .urgency-low { background: var(--accent-orange); }
+        .urgency-high { background: var(--accent-red); }
+
+        @keyframes pulse {
+            0%, 100% { opacity: 1; }
+            50% { opacity: 0.5; }
+        }
+
+        .updating {
+            animation: pulse 1s infinite;
+        }
+    </style>
+</head>
+<body>
+    <header class="header">
+        <h1>🧠 GOAP Debug Visualizer</h1>
+        <div class="header-controls">
+            <span id="turn-display">Turn 0</span>
+            <span id="status-badge" class="status-badge status-disconnected">Disconnected</span>
+            <button class="btn" onclick="refreshData()">↻ Refresh</button>
+            <button class="btn btn-primary" id="auto-refresh-btn" onclick="toggleAutoRefresh()">▶ Auto</button>
+        </div>
+    </header>
+
+    <main class="main-content">
+        <!-- Left Panel: Agent List -->
+        <div class="panel">
+            <div class="panel-header">
+                <h2>Agents</h2>
+            </div>
+            <div id="agent-list" class="agent-list">
+                <div class="loading">Loading...</div>
+            </div>
+        </div>
+
+        <!-- Center Panel: Plan Visualization -->
+        <div class="center-panel">
+            <div class="plan-view" id="plan-view">
+                <div class="loading">Select an agent to view GOAP details</div>
+            </div>
+        </div>
+
+        <!-- Right Panel: Details -->
+        <div class="panel">
+            <div class="panel-header">
+                <h2>Details</h2>
+            </div>
+            <div id="details-panel">
+                <div class="loading">Select an agent</div>
+            </div>
+        </div>
+    </main>
+
+    <script>
+        const API_BASE = 'http://localhost:8000/api';
+        let selectedAgentId = null;
+        let allAgentsData = [];
+        let autoRefreshInterval = null;
+        let goalsChart = null;
+
+        // Initialize
+        document.addEventListener('DOMContentLoaded', () => {
+            refreshData();
+        });
+
+        async function refreshData() {
+            try {
+                const response = await fetch(`${API_BASE}/goap/debug`);
+                if (!response.ok) throw new Error('API error');
+
+                const data = await response.json();
+                allAgentsData = data.agents;
+
+                document.getElementById('turn-display').textContent = `Turn ${data.current_turn}`;
+                document.getElementById('status-badge').className = 'status-badge status-connected';
+                document.getElementById('status-badge').textContent = data.is_night ? '🌙 Night' : '☀️ Connected';
+
+                renderAgentList();
+
+                if (selectedAgentId) {
+                    const agent = allAgentsData.find(a => a.agent_id === selectedAgentId);
+                    if (agent) {
+                        renderAgentDetails(agent);
+                    }
+                }
+            } catch (error) {
+                console.error('Failed to fetch data:', error);
+                document.getElementById('status-badge').className = 'status-badge status-disconnected';
+                document.getElementById('status-badge').textContent = 'Disconnected';
+            }
+        }
+
+        function renderAgentList() {
+            const container = document.getElementById('agent-list');
+
+            if (allAgentsData.length === 0) {
+                container.innerHTML = '<div class="loading">No agents found</div>';
+                return;
+            }
+
+            container.innerHTML = allAgentsData.map(agent => `
+                <div class="agent-item ${agent.agent_id === selectedAgentId ? 'selected' : ''}"
+                     onclick="selectAgent('${agent.agent_id}')">
+                    <div class="agent-name">${agent.agent_name}</div>
+                    <div class="agent-action">${agent.selected_action || 'No action'}</div>
+                    <div class="agent-goal">${agent.current_plan ? '🎯 ' + agent.current_plan.goal_name : '(reactive)'}</div>
+                </div>
+            `).join('');
+        }
+
+        function selectAgent(agentId) {
+            selectedAgentId = agentId;
+            renderAgentList();
+
+            const agent = allAgentsData.find(a => a.agent_id === agentId);
+            if (agent) {
+                renderAgentDetails(agent);
+            }
+        }
+
+        function renderAgentDetails(agent) {
+            renderPlanView(agent);
+            renderDetailsPanel(agent);
+        }
+
+        function getUrgencyClass(urgency) {
+            if (urgency <= 0) return 'urgency-none';
+            if (urgency <= 1) return 'urgency-low';
+            return 'urgency-high';
+        }
+
+        function renderPlanView(agent) {
+            const container = document.getElementById('plan-view');
+            const ws = agent.world_state;
+            const plan = agent.current_plan;
+
+            container.innerHTML = `
+                <div class="plan-header">
+                    <h2>${agent.agent_name}</h2>
+                    ${plan ? `<span class="plan-goal-badge">🎯 ${plan.goal_name}</span>` : ''}
+                </div>
+
+                <div class="world-state-grid">
+                    <div class="state-card bar-thirst">
+                        <div class="label">Thirst</div>
+                        <div class="value">${Math.round(ws.vitals.thirst * 100)}%
+                            <span class="urgency-indicator ${getUrgencyClass(ws.urgencies.thirst)}"></span>
+                        </div>
+                        <div class="bar"><div class="bar-fill" style="width: ${ws.vitals.thirst * 100}%"></div></div>
+                    </div>
+                    <div class="state-card bar-hunger">
+                        <div class="label">Hunger</div>
+                        <div class="value">${Math.round(ws.vitals.hunger * 100)}%
+                            <span class="urgency-indicator ${getUrgencyClass(ws.urgencies.hunger)}"></span>
+                        </div>
+                        <div class="bar"><div class="bar-fill" style="width: ${ws.vitals.hunger * 100}%"></div></div>
+                    </div>
+                    <div class="state-card bar-heat">
+                        <div class="label">Heat</div>
+                        <div class="value">${Math.round(ws.vitals.heat * 100)}%
+                            <span class="urgency-indicator ${getUrgencyClass(ws.urgencies.heat)}"></span>
+                        </div>
+                        <div class="bar"><div class="bar-fill" style="width: ${ws.vitals.heat * 100}%"></div></div>
+                    </div>
+                    <div class="state-card bar-energy">
+                        <div class="label">Energy</div>
+                        <div class="value">${Math.round(ws.vitals.energy * 100)}%
+                            <span class="urgency-indicator ${getUrgencyClass(ws.urgencies.energy)}"></span>
+                        </div>
+                        <div class="bar"><div class="bar-fill" style="width: ${ws.vitals.energy * 100}%"></div></div>
+                    </div>
+                </div>
+
+                <div class="plan-visualization">
+                    <h3>Current Plan</h3>
+                    ${plan && plan.actions.length > 0 ? `
+                        <div class="plan-steps">
+                            ${plan.actions.map((action, i) => `
+                                <div class="plan-step">
+                                    <div class="step-node ${i === 0 ? 'current' : ''}">${action}</div>
+                                    ${i < plan.actions.length - 1 ? '<span class="step-arrow">→</span>' : ''}
+                                </div>
+                            `).join('')}
+                            <span class="step-arrow">→</span>
+                            <div class="goal-result">✓ ${plan.goal_name}</div>
+                        </div>
+                        <div style="margin-top: 12px; font-size: 13px; color: var(--text-muted);">
+                            Total Cost: ${plan.total_cost.toFixed(1)} | Steps: ${plan.plan_length}
+                        </div>
+                    ` : `
+                        <div class="no-plan">
+                            <p style="font-size: 16px; margin-bottom: 8px;">No plan - using reactive selection</p>
+                            <p>Selected: <strong>${agent.selected_action || 'None'}</strong></p>
+                        </div>
+                    `}
+                </div>
+
+                <div class="goals-chart-container">
+                    <h3>Goal Priorities</h3>
+                    <div class="chart-wrapper">
+                        <canvas id="goals-chart"></canvas>
+                    </div>
+                </div>
+            `;
+
+            renderGoalsChart(agent);
+        }
+
+        function renderGoalsChart(agent) {
+            const ctx = document.getElementById('goals-chart');
+            if (!ctx) return;
+
+            // Sort goals by priority
+            const sortedGoals = [...agent.goals].sort((a, b) => b.priority - a.priority);
+            const topGoals = sortedGoals.slice(0, 10);
+
+            if (goalsChart) {
+                goalsChart.destroy();
+            }
+
+            goalsChart = new Chart(ctx, {
+                type: 'bar',
+                data: {
+                    labels: topGoals.map(g => g.name),
+                    datasets: [{
+                        label: 'Priority',
+                        data: topGoals.map(g => g.priority),
+                        backgroundColor: topGoals.map(g => {
+                            if (g.is_selected) return 'rgba(163, 113, 247, 0.8)';
+                            if (g.is_satisfied) return 'rgba(63, 185, 80, 0.5)';
+                            if (g.priority > 0) return 'rgba(88, 166, 255, 0.7)';
+                            return 'rgba(110, 118, 129, 0.3)';
+                        }),
+                        borderColor: topGoals.map(g => {
+                            if (g.is_selected) return '#a371f7';
+                            if (g.is_satisfied) return '#3fb950';
+                            if (g.priority > 0) return '#58a6ff';
+                            return '#6e7681';
+                        }),
+                        borderWidth: 2,
+                    }]
+                },
+                options: {
+                    indexAxis: 'y',
+                    responsive: true,
+                    maintainAspectRatio: false,
+                    plugins: {
+                        legend: { display: false },
+                    },
+                    scales: {
+                        x: {
+                            beginAtZero: true,
+                            grid: { color: '#30363d' },
+                            ticks: { color: '#8b949e' },
+                        },
+                        y: {
+                            grid: { display: false },
+                            ticks: {
+                                color: '#e6edf3',
+                                font: { family: 'IBM Plex Mono', size: 11 }
+                            },
+                        }
+                    }
+                }
+            });
+        }
+
+        function renderDetailsPanel(agent) {
+            const container = document.getElementById('details-panel');
+            const ws = agent.world_state;
+
+            const validActions = agent.actions.filter(a => a.is_valid);
+            const inPlanActions = agent.actions.filter(a => a.is_in_plan).sort((a, b) => a.plan_order - b.plan_order);
+
+            container.innerHTML = `
+                <div class="detail-section">
+                    <h3>Inventory</h3>
+                    <div class="inventory-grid">
+                        <div class="inv-item"><span class="icon">💧</span> Water <span class="count">${ws.inventory.water}</span></div>
+                        <div class="inv-item"><span class="icon">🍖</span> Meat <span class="count">${ws.inventory.meat}</span></div>
+                        <div class="inv-item"><span class="icon">🫐</span> Berries <span class="count">${ws.inventory.berries}</span></div>
+                        <div class="inv-item"><span class="icon">🪵</span> Wood <span class="count">${ws.inventory.wood}</span></div>
+                        <div class="inv-item"><span class="icon">🥩</span> Hide <span class="count">${ws.inventory.hide}</span></div>
+                        <div class="inv-item"><span class="icon">📦</span> Space <span class="count">${ws.inventory.space}</span></div>
+                    </div>
+                </div>
+
+                <div class="detail-section">
+                    <h3>Economy</h3>
+                    <div class="detail-grid">
+                        <div class="detail-item">
+                            <span class="label">Money</span>
+                            <span class="value" style="color: var(--accent-orange)">${ws.economy.money}c</span>
+                        </div>
+                        <div class="detail-item">
+                            <span class="label">Wealthy</span>
+                            <span class="value">${ws.economy.is_wealthy ? '✓' : '✗'}</span>
+                        </div>
+                    </div>
+                </div>
+
+                <div class="detail-section">
+                    <h3>Market Access</h3>
+                    <div class="detail-grid">
+                        <div class="detail-item">
+                            <span class="label">Buy Water</span>
+                            <span class="value">${ws.market.can_buy_water ? '✓' : '✗'}</span>
+                        </div>
+                        <div class="detail-item">
+                            <span class="label">Buy Food</span>
+                            <span class="value">${ws.market.can_buy_food ? '✓' : '✗'}</span>
+                        </div>
+                        <div class="detail-item">
+                            <span class="label">Buy Wood</span>
+                            <span class="value">${ws.market.can_buy_wood ? '✓' : '✗'}</span>
+                        </div>
+                    </div>
+                </div>
+
+                <div class="detail-section">
+                    <h3>Actions (${validActions.length} valid)</h3>
+                    <div class="action-list">
+                        ${agent.actions.map(action => `
+                            <div class="action-item ${action.is_valid ? 'valid' : 'invalid'} ${action.is_in_plan ? 'in-plan' : ''}">
+                                ${action.is_in_plan ? `<span class="action-order">${action.plan_order + 1}</span>` : ''}
+                                <span class="action-name">${action.name}</span>
+                                <span class="action-cost">${action.cost >= 0 ? action.cost.toFixed(1) : '∞'}</span>
+                            </div>
+                        `).join('')}
+                    </div>
+                </div>
+            `;
+        }
+
+        function toggleAutoRefresh() {
+            const btn = document.getElementById('auto-refresh-btn');
+
+            if (autoRefreshInterval) {
+                clearInterval(autoRefreshInterval);
+                autoRefreshInterval = null;
+                btn.textContent = '▶ Auto';
+                btn.classList.remove('btn-primary');
+            } else {
+                autoRefreshInterval = setInterval(refreshData, 500);
+                btn.textContent = '⏸ Stop';
+                btn.classList.add('btn-primary');
+            }
+        }
+    </script>
+</body>
+</html>
+

web_frontend/index.html

@@ -135,6 +135,7 @@
                         <button class="tab-btn" data-tab="resources">Resources</button>
                         <button class="tab-btn" data-tab="market">Market</button>
                         <button class="tab-btn" data-tab="agents">Agents</button>
+                        <button class="tab-btn" data-tab="goap">🧠 GOAP</button>
                     </div>
                 </div>
                 <div class="stats-header-right">
@@ -240,8 +241,54 @@
                         </div>
                     </div>
                 </div>
+                <!-- GOAP Tab -->
+                <div id="tab-goap" class="tab-panel">
+                    <div class="goap-container">
+                        <div class="goap-header">
+                            <h3>Goal-Oriented Action Planning</h3>
+                            <p class="goap-subtitle">Real-time visualization of agent decision-making</p>
+                        </div>
+                        <div class="goap-grid">
+                            <div class="goap-panel goap-agents-panel">
+                                <h4>Agents</h4>
+                                <div id="goap-agent-list" class="goap-agent-list">
+                                    <p class="loading-text">Loading agents...</p>
+                                </div>
+                            </div>
+                            <div class="goap-panel goap-plan-panel">
+                                <h4>Current Plan</h4>
+                                <div id="goap-plan-view" class="goap-plan-view">
+                                    <p class="no-selection-text">Select an agent to view their GOAP plan</p>
+                                </div>
+                            </div>
+                            <div class="goap-panel goap-goals-panel">
+                                <h4>Goal Priorities</h4>
+                                <div class="chart-wrapper">
+                                    <canvas id="chart-goap-goals"></canvas>
+                                </div>
+                            </div>
+                            <div class="goap-panel goap-actions-panel">
+                                <h4>Available Actions</h4>
+                                <div id="goap-actions-list" class="goap-actions-list">
+                                    <p class="no-selection-text">Select an agent</p>
+                                </div>
+                            </div>
+                        </div>
+                    </div>
+                </div>
             </div>
             <div class="stats-footer">
+                <div class="controls">
+                    <button id="btn-initialize-stats" class="btn btn-secondary" title="Reset Simulation">
+                        <span class="btn-icon">⟳</span> Reset
+                    </button>
+                    <button id="btn-step-stats" class="btn btn-primary" title="Advance one turn">
+                        <span class="btn-icon">▶</span> Step
+                    </button>
+                    <button id="btn-auto-stats" class="btn btn-toggle" title="Toggle auto mode">
+                        <span class="btn-icon">⏯</span> Auto
+                    </button>
+                </div>
                 <div class="stats-summary-bar">
                     <div class="summary-item">
                         <span class="summary-label">Turn</span>
@@ -268,6 +315,11 @@
                         <span class="summary-value" id="stats-gini">0.00</span>
                     </div>
                 </div>
+                <div class="speed-control">
+                    <label for="speed-slider-stats">Speed</label>
+                    <input type="range" id="speed-slider-stats" min="50" max="1000" value="150" step="50">
+                    <span id="speed-display-stats">150ms</span>
+                </div>
             </div>
         </div>
     </div>

web_frontend/src/api.js

@@ -124,6 +124,21 @@ class SimulationAPI {
     async getLogs(limit = 10) {
         return await this.request(`/api/logs?limit=${limit}`);
     }
+
+    // GOAP: Get debug info for all agents
+    async getGOAPDebug() {
+        return await this.request('/api/goap/debug');
+    }
+
+    // GOAP: Get debug info for specific agent
+    async getAgentGOAPDebug(agentId) {
+        return await this.request(`/api/goap/debug/${agentId}`);
+    }
+
+    // Generic GET helper (for compatibility)
+    async get(endpoint) {
+        return await this.request(`/api${endpoint}`);
+    }
 }
 
 // Export singleton instance

web_frontend/src/scenes/GameScene.js

@@ -127,7 +127,22 @@ export default class GameScene extends Phaser.Scene {
             statsGold: document.getElementById('stats-gold'),
             statsAvgWealth: document.getElementById('stats-avg-wealth'),
             statsGini: document.getElementById('stats-gini'),
+            // GOAP elements
+            goapAgentList: document.getElementById('goap-agent-list'),
+            goapPlanView: document.getElementById('goap-plan-view'),
+            goapActionsList: document.getElementById('goap-actions-list'),
+            chartGoapGoals: document.getElementById('chart-goap-goals'),
+            // Stats screen controls (duplicated for stats page)
+            btnStepStats: document.getElementById('btn-step-stats'),
+            btnAutoStats: document.getElementById('btn-auto-stats'),
+            btnInitializeStats: document.getElementById('btn-initialize-stats'),
+            speedSliderStats: document.getElementById('speed-slider-stats'),
+            speedDisplayStats: document.getElementById('speed-display-stats'),
         };
+
+        // GOAP state
+        this.goapData = null;
+        this.selectedGoapAgentId = null;
     }
 
     cleanup() {
@@ -163,6 +178,21 @@ export default class GameScene extends Phaser.Scene {
             btnCloseStats.removeEventListener('click', this.boundHandlers.closeStats);
         }
 
+        // Stats screen controls cleanup
+        const { btnStepStats, btnAutoStats, btnInitializeStats, speedSliderStats } = this.domCache;
+        if (btnStepStats && this.boundHandlers.step) {
+            btnStepStats.removeEventListener('click', this.boundHandlers.step);
+        }
+        if (btnAutoStats && this.boundHandlers.auto) {
+            btnAutoStats.removeEventListener('click', this.boundHandlers.auto);
+        }
+        if (btnInitializeStats && this.boundHandlers.init) {
+            btnInitializeStats.removeEventListener('click', this.boundHandlers.init);
+        }
+        if (speedSliderStats && this.boundHandlers.speedStats) {
+            speedSliderStats.removeEventListener('input', this.boundHandlers.speedStats);
+        }
+
         // Destroy charts
         Object.values(this.charts).forEach(chart => chart?.destroy());
         this.charts = {};
@@ -277,19 +307,34 @@ export default class GameScene extends Phaser.Scene {
 
     setupUIControls() {
         const { btnStep, btnAuto, btnInitialize, btnStats, btnCloseStats, speedSlider, speedDisplay, tabButtons } = this.domCache;
+        const { btnStepStats, btnAutoStats, btnInitializeStats, speedSliderStats, speedDisplayStats } = this.domCache;
 
         // Create bound handlers for later cleanup
         this.boundHandlers.step = () => this.handleStep();
         this.boundHandlers.auto = () => this.toggleAutoMode();
         this.boundHandlers.init = () => this.handleInitialize();
+
+        // Speed handler that syncs both sliders
         this.boundHandlers.speed = (e) => {
             this.autoSpeed = parseInt(e.target.value);
             if (speedDisplay) speedDisplay.textContent = `${this.autoSpeed}ms`;
+            if (speedDisplayStats) speedDisplayStats.textContent = `${this.autoSpeed}ms`;
+            if (speedSliderStats) speedSliderStats.value = this.autoSpeed;
             if (this.isAutoMode) this.restartAutoMode();
         };
+
+        this.boundHandlers.speedStats = (e) => {
+            this.autoSpeed = parseInt(e.target.value);
+            if (speedDisplay) speedDisplay.textContent = `${this.autoSpeed}ms`;
+            if (speedDisplayStats) speedDisplayStats.textContent = `${this.autoSpeed}ms`;
+            if (speedSlider) speedSlider.value = this.autoSpeed;
+            if (this.isAutoMode) this.restartAutoMode();
+        };
+
         this.boundHandlers.openStats = () => this.showStatsScreen();
         this.boundHandlers.closeStats = () => this.hideStatsScreen();
 
+        // Main controls
         if (btnStep) btnStep.addEventListener('click', this.boundHandlers.step);
         if (btnAuto) btnAuto.addEventListener('click', this.boundHandlers.auto);
         if (btnInitialize) btnInitialize.addEventListener('click', this.boundHandlers.init);
@@ -297,6 +342,12 @@ export default class GameScene extends Phaser.Scene {
         if (btnStats) btnStats.addEventListener('click', this.boundHandlers.openStats);
         if (btnCloseStats) btnCloseStats.addEventListener('click', this.boundHandlers.closeStats);
 
+        // Stats screen controls (same handlers)
+        if (btnStepStats) btnStepStats.addEventListener('click', this.boundHandlers.step);
+        if (btnAutoStats) btnAutoStats.addEventListener('click', this.boundHandlers.auto);
+        if (btnInitializeStats) btnInitializeStats.addEventListener('click', this.boundHandlers.init);
+        if (speedSliderStats) speedSliderStats.addEventListener('input', this.boundHandlers.speedStats);
+
         // Tab switching
         tabButtons?.forEach(btn => {
             btn.addEventListener('click', (e) => this.switchTab(e.target.dataset.tab));
@@ -371,15 +422,19 @@ export default class GameScene extends Phaser.Scene {
 
     toggleAutoMode() {
         this.isAutoMode = !this.isAutoMode;
-        const { btnAuto, btnStep } = this.domCache;
+        const { btnAuto, btnStep, btnAutoStats, btnStepStats } = this.domCache;
 
         if (this.isAutoMode) {
             btnAuto?.classList.add('active');
+            btnAutoStats?.classList.add('active');
             btnStep?.setAttribute('disabled', 'true');
+            btnStepStats?.setAttribute('disabled', 'true');
             this.startAutoMode();
         } else {
             btnAuto?.classList.remove('active');
+            btnAutoStats?.classList.remove('active');
             btnStep?.removeAttribute('disabled');
+            btnStepStats?.removeAttribute('disabled');
             this.stopAutoMode();
         }
     }
@@ -741,6 +796,15 @@ export default class GameScene extends Phaser.Scene {
                 <span class="action-label">Current Action</span>
                 <div>${actionData.icon} ${action.message || actionData.verb}</div>
             </div>
+            <div class="agent-goap-info" id="agent-goap-section" data-agent-id="${agentData.id}">
+                <h5 class="subsection-title" style="display: flex; align-items: center; gap: 6px;">
+                    🧠 GOAP Plan
+                    <button class="btn-mini" onclick="window.villsimGame.scene.scenes[1].loadAgentGOAP('${agentData.id}')" style="font-size: 0.6rem; padding: 2px 6px;">↻</button>
+                </h5>
+                <div id="agent-goap-content" style="font-size: 0.75rem; color: var(--text-muted);">
+                    Click ↻ to load GOAP info
+                </div>
+            </div>
             <h5 class="subsection-title">Personal Log</h5>
             <div class="agent-log">
                 ${renderActionLog()}
@@ -1023,6 +1087,7 @@ export default class GameScene extends Phaser.Scene {
             case 'resources': this.renderResourceCharts(); break;
             case 'market': this.renderMarketCharts(); break;
             case 'agents': this.renderAgentStatsCharts(); break;
+            case 'goap': this.fetchAndRenderGOAP(); break;
         }
     }
 
@@ -1627,6 +1692,291 @@ export default class GameScene extends Phaser.Scene {
         };
     }
 
+    // =================================
+    // GOAP Visualization Methods
+    // =================================
+
+    async loadAgentGOAP(agentId) {
+        const contentEl = document.getElementById('agent-goap-content');
+        if (!contentEl) return;
+
+        contentEl.innerHTML = '<span style="color: var(--text-muted);">Loading...</span>';
+
+        try {
+            const data = await api.getAgentGOAPDebug(agentId);
+            const plan = data.current_plan;
+
+            if (plan && plan.actions.length > 0) {
+                contentEl.innerHTML = `
+                    <div style="margin-bottom: 4px;">
+                        <strong style="color: var(--accent-sapphire);">Goal:</strong> ${plan.goal_name}
+                    </div>
+                    <div style="font-family: var(--font-mono); font-size: 0.7rem;">
+                        ${plan.actions.map((a, i) =>
+                            `<span style="${i === 0 ? 'color: var(--accent-emerald);' : ''}">${a}</span>`
+                        ).join(' → ')}
+                    </div>
+                    <div style="margin-top: 4px; color: var(--text-muted); font-size: 0.65rem;">
+                        Cost: ${plan.total_cost.toFixed(1)} | Steps: ${plan.plan_length}
+                    </div>
+                `;
+            } else {
+                contentEl.innerHTML = `
+                    <div style="color: var(--text-muted);">
+                        No plan (reactive mode)<br>
+                        <span style="color: var(--text-primary);">${data.selected_action || 'No action'}</span>
+                    </div>
+                `;
+            }
+        } catch (error) {
+            console.error('Failed to load GOAP info:', error);
+            contentEl.innerHTML = '<span style="color: var(--accent-ruby);">Failed to load</span>';
+        }
+    }
+
+    async fetchAndRenderGOAP() {
+        try {
+            const response = await api.get('/goap/debug');
+            this.goapData = response;
+            this.renderGOAPAgentList();
+
+            // If we have a selected agent, render their details
+            if (this.selectedGoapAgentId) {
+                const agent = this.goapData.agents.find(a => a.agent_id === this.selectedGoapAgentId);
+                if (agent) {
+                    this.renderGOAPAgentDetails(agent);
+                }
+            }
+        } catch (error) {
+            console.error('Failed to fetch GOAP data:', error);
+            const { goapAgentList } = this.domCache;
+            if (goapAgentList) {
+                goapAgentList.innerHTML = '<p class="loading-text">Failed to load GOAP data. Make sure the server is running.</p>';
+            }
+        }
+    }
+
+    renderGOAPAgentList() {
+        const { goapAgentList } = this.domCache;
+        if (!goapAgentList || !this.goapData) return;
+
+        if (this.goapData.agents.length === 0) {
+            goapAgentList.innerHTML = '<p class="loading-text">No agents found</p>';
+            return;
+        }
+
+        goapAgentList.innerHTML = this.goapData.agents.map(agent => `
+            <div class="goap-agent-item ${agent.agent_id === this.selectedGoapAgentId ? 'selected' : ''}"
+                 data-agent-id="${agent.agent_id}">
+                <div class="agent-name">${agent.agent_name}</div>
+                <div class="agent-action">${agent.selected_action || 'No action'}</div>
+                <div class="agent-goal">${agent.current_plan ? '🎯 ' + agent.current_plan.goal_name : '(reactive)'}</div>
+            </div>
+        `).join('');
+
+        // Add click handlers
+        goapAgentList.querySelectorAll('.goap-agent-item').forEach(item => {
+            item.addEventListener('click', () => {
+                this.selectGoapAgent(item.dataset.agentId);
+            });
+        });
+    }
+
+    selectGoapAgent(agentId) {
+        this.selectedGoapAgentId = agentId;
+
+        // Update selection styling
+        const { goapAgentList } = this.domCache;
+        if (goapAgentList) {
+            goapAgentList.querySelectorAll('.goap-agent-item').forEach(item => {
+                item.classList.toggle('selected', item.dataset.agentId === agentId);
+            });
+        }
+
+        // Render details
+        if (this.goapData) {
+            const agent = this.goapData.agents.find(a => a.agent_id === agentId);
+            if (agent) {
+                this.renderGOAPAgentDetails(agent);
+            }
+        }
+    }
+
+    renderGOAPAgentDetails(agent) {
+        this.renderGOAPPlanView(agent);
+        this.renderGOAPActionsList(agent);
+        this.renderGOAPGoalsChart(agent);
+    }
+
+    getUrgencyClass(urgency) {
+        if (urgency <= 0) return 'none';
+        if (urgency <= 1) return 'low';
+        return 'high';
+    }
+
+    renderGOAPPlanView(agent) {
+        const { goapPlanView } = this.domCache;
+        if (!goapPlanView) return;
+
+        const ws = agent.world_state;
+        const plan = agent.current_plan;
+
+        goapPlanView.innerHTML = `
+            <div class="goap-world-state">
+                <div class="goap-stat-card thirst">
+                    <div class="label">Thirst</div>
+                    <div class="value">${Math.round(ws.vitals.thirst * 100)}%<span class="goap-urgency ${this.getUrgencyClass(ws.urgencies.thirst)}"></span></div>
+                    <div class="bar"><div class="bar-fill" style="width: ${ws.vitals.thirst * 100}%"></div></div>
+                </div>
+                <div class="goap-stat-card hunger">
+                    <div class="label">Hunger</div>
+                    <div class="value">${Math.round(ws.vitals.hunger * 100)}%<span class="goap-urgency ${this.getUrgencyClass(ws.urgencies.hunger)}"></span></div>
+                    <div class="bar"><div class="bar-fill" style="width: ${ws.vitals.hunger * 100}%"></div></div>
+                </div>
+                <div class="goap-stat-card heat">
+                    <div class="label">Heat</div>
+                    <div class="value">${Math.round(ws.vitals.heat * 100)}%<span class="goap-urgency ${this.getUrgencyClass(ws.urgencies.heat)}"></span></div>
+                    <div class="bar"><div class="bar-fill" style="width: ${ws.vitals.heat * 100}%"></div></div>
+                </div>
+                <div class="goap-stat-card energy">
+                    <div class="label">Energy</div>
+                    <div class="value">${Math.round(ws.vitals.energy * 100)}%<span class="goap-urgency ${this.getUrgencyClass(ws.urgencies.energy)}"></span></div>
+                    <div class="bar"><div class="bar-fill" style="width: ${ws.vitals.energy * 100}%"></div></div>
+                </div>
+            </div>
+
+            <div class="goap-plan-steps">
+                <h5>Current Plan</h5>
+                ${plan && plan.actions.length > 0 ? `
+                    <div class="goap-plan-flow">
+                        ${plan.actions.map((action, i) => `
+                            <span class="goap-step-node ${i === 0 ? 'current' : ''}">${action}</span>
+                            ${i < plan.actions.length - 1 ? '<span class="goap-step-arrow">→</span>' : ''}
+                        `).join('')}
+                        <span class="goap-step-arrow">→</span>
+                        <span class="goap-goal-result">✓ ${plan.goal_name}</span>
+                    </div>
+                    <div style="margin-top: 8px; font-size: 0.75rem; color: var(--text-muted);">
+                        Total Cost: ${plan.total_cost.toFixed(1)} | Steps: ${plan.plan_length}
+                    </div>
+                ` : `
+                    <div style="color: var(--text-muted); font-size: 0.85rem;">
+                        No plan - using reactive selection<br>
+                        Selected: <strong>${agent.selected_action || 'None'}</strong>
+                    </div>
+                `}
+            </div>
+
+            <div style="margin-top: 12px;">
+                <h5 style="font-size: 0.7rem; color: var(--text-muted); margin-bottom: 8px;">INVENTORY</h5>
+                <div class="goap-inventory">
+                    <div class="goap-inv-item">💧<span class="count">${ws.inventory.water}</span></div>
+                    <div class="goap-inv-item">🍖<span class="count">${ws.inventory.meat}</span></div>
+                    <div class="goap-inv-item">🫐<span class="count">${ws.inventory.berries}</span></div>
+                    <div class="goap-inv-item">🪵<span class="count">${ws.inventory.wood}</span></div>
+                    <div class="goap-inv-item">🥩<span class="count">${ws.inventory.hide}</span></div>
+                    <div class="goap-inv-item">📦<span class="count">${ws.inventory.space}</span></div>
+                </div>
+            </div>
+
+            <div style="margin-top: 12px; display: flex; gap: 16px; font-size: 0.8rem;">
+                <span style="color: var(--accent-gold);">💰 ${ws.economy.money}c</span>
+                <span style="color: var(--text-muted);">Wealthy: ${ws.economy.is_wealthy ? '✓' : '✗'}</span>
+            </div>
+        `;
+    }
+
+    renderGOAPActionsList(agent) {
+        const { goapActionsList } = this.domCache;
+        if (!goapActionsList) return;
+
+        // Sort: plan actions first, then valid, then invalid
+        const sortedActions = [...agent.actions].sort((a, b) => {
+            if (a.is_in_plan && !b.is_in_plan) return -1;
+            if (!a.is_in_plan && b.is_in_plan) return 1;
+            if (a.is_in_plan && b.is_in_plan) return a.plan_order - b.plan_order;
+            if (a.is_valid && !b.is_valid) return -1;
+            if (!a.is_valid && b.is_valid) return 1;
+            return (a.cost || 999) - (b.cost || 999);
+        });
+
+        goapActionsList.innerHTML = sortedActions.map(action => `
+            <div class="goap-action-item ${action.is_valid ? 'valid' : 'invalid'} ${action.is_in_plan ? 'in-plan' : ''}">
+                ${action.is_in_plan ? `<span class="action-order">${action.plan_order + 1}</span>` : ''}
+                <span class="action-name">${action.name}</span>
+                <span class="action-cost">${action.cost >= 0 ? action.cost.toFixed(1) : '∞'}</span>
+            </div>
+        `).join('');
+    }
+
+    renderGOAPGoalsChart(agent) {
+        const { chartGoapGoals } = this.domCache;
+        if (!chartGoapGoals) return;
+
+        // Sort goals by priority and take top 10
+        const sortedGoals = [...agent.goals]
+            .sort((a, b) => b.priority - a.priority)
+            .slice(0, 10);
+
+        // Destroy existing chart
+        if (this.charts.goapGoals) {
+            this.charts.goapGoals.destroy();
+        }
+
+        this.charts.goapGoals = new Chart(chartGoapGoals, {
+            type: 'bar',
+            data: {
+                labels: sortedGoals.map(g => g.name),
+                datasets: [{
+                    label: 'Priority',
+                    data: sortedGoals.map(g => g.priority),
+                    backgroundColor: sortedGoals.map(g => {
+                        if (g.is_selected) return 'rgba(139, 111, 192, 0.8)';
+                        if (g.is_satisfied) return 'rgba(74, 156, 109, 0.5)';
+                        if (g.priority > 0) return 'rgba(90, 140, 200, 0.7)';
+                        return 'rgba(107, 101, 96, 0.3)';
+                    }),
+                    borderColor: sortedGoals.map(g => {
+                        if (g.is_selected) return '#8b6fc0';
+                        if (g.is_satisfied) return '#4a9c6d';
+                        if (g.priority > 0) return '#5a8cc8';
+                        return '#6b6560';
+                    }),
+                    borderWidth: 2,
+                }]
+            },
+            options: {
+                indexAxis: 'y',
+                responsive: true,
+                maintainAspectRatio: false,
+                animation: false,
+                plugins: {
+                    legend: { display: false },
+                    title: {
+                        display: true,
+                        text: 'Goal Priorities',
+                        color: '#e8e4dc',
+                        font: { family: "'Crimson Pro', serif", size: 14 },
+                    },
+                },
+                scales: {
+                    x: {
+                        beginAtZero: true,
+                        grid: { color: 'rgba(58, 67, 89, 0.3)' },
+                        ticks: { color: '#6b6560' },
+                    },
+                    y: {
+                        grid: { display: false },
+                        ticks: {
+                            color: '#e8e4dc',
+                            font: { family: "'JetBrains Mono', monospace", size: 10 }
+                        },
+                    }
+                }
+            }
+        });
+    }
+
     update(time, delta) {
         // Minimal update loop - no heavy operations here
     }

web_frontend/styles.css

@@ -613,7 +613,8 @@ body {
     font-weight: 500;
 }
 
-#speed-slider {
+#speed-slider,
+#speed-slider-stats {
     width: 120px;
     height: 4px;
     -webkit-appearance: none;
@@ -623,7 +624,8 @@ body {
     cursor: pointer;
 }
 
-#speed-slider::-webkit-slider-thumb {
+#speed-slider::-webkit-slider-thumb,
+#speed-slider-stats::-webkit-slider-thumb {
     -webkit-appearance: none;
     appearance: none;
     width: 14px;
@@ -633,7 +635,8 @@ body {
     cursor: pointer;
 }
 
-#speed-display {
+#speed-display,
+#speed-display-stats {
     font-family: var(--font-mono);
     min-width: 50px;
 }
@@ -939,16 +942,21 @@ body {
 
 /* Stats Footer */
 .stats-footer {
+    display: flex;
+    align-items: center;
+    justify-content: space-between;
     padding: var(--space-sm) var(--space-lg);
     background: var(--bg-primary);
     border-top: 1px solid var(--border-color);
     flex-shrink: 0;
+    height: 56px;
 }
 
 .stats-summary-bar {
     display: flex;
     justify-content: center;
     gap: var(--space-xl);
+    flex: 1;
 }
 
 .summary-item {
@@ -1049,5 +1057,375 @@ body {
         flex-wrap: wrap;
         gap: var(--space-md);
     }
+
+    .stats-footer {
+        flex-wrap: wrap;
+        height: auto;
+        gap: var(--space-sm);
+        padding: var(--space-sm);
+    }
+
+    .stats-footer .controls {
+        order: 1;
+        width: auto;
+    }
+
+    .stats-footer .stats-summary-bar {
+        order: 2;
+        width: 100%;
+    }
+
+    .stats-footer .speed-control {
+        order: 3;
+        width: auto;
+    }
+}
+
+/* =================================
+   GOAP Visualization Styles
+   ================================= */
+
+.goap-container {
+    padding: var(--space-lg);
+    height: 100%;
+    display: flex;
+    flex-direction: column;
+}
+
+.goap-header {
+    margin-bottom: var(--space-lg);
+}
+
+.goap-header h3 {
+    font-size: 1.5rem;
+    color: var(--accent-sapphire);
+    margin-bottom: var(--space-xs);
+}
+
+.goap-subtitle {
+    font-size: 0.85rem;
+    color: var(--text-muted);
+}
+
+.goap-grid {
+    display: grid;
+    grid-template-columns: 250px 1fr 300px;
+    grid-template-rows: 1fr 1fr;
+    gap: var(--space-md);
+    flex: 1;
+    min-height: 0;
+}
+
+.goap-panel {
+    background: var(--bg-secondary);
+    border: 1px solid var(--border-color);
+    border-radius: var(--radius-md);
+    padding: var(--space-md);
+    display: flex;
+    flex-direction: column;
+    overflow: hidden;
+}
+
+.goap-panel h4 {
+    font-size: 0.8rem;
+    text-transform: uppercase;
+    letter-spacing: 0.5px;
+    color: var(--text-muted);
+    margin-bottom: var(--space-md);
+    padding-bottom: var(--space-sm);
+    border-bottom: 1px solid var(--border-color);
+}
+
+.goap-agents-panel {
+    grid-row: span 2;
+}
+
+.goap-plan-panel {
+    grid-column: 2;
+}
+
+.goap-goals-panel {
+    grid-column: 3;
+    grid-row: span 2;
+}
+
+.goap-actions-panel {
+    grid-column: 2;
+}
+
+.goap-agent-list {
+    flex: 1;
+    overflow-y: auto;
+}
+
+.goap-agent-item {
+    padding: var(--space-sm) var(--space-md);
+    border-radius: var(--radius-sm);
+    margin-bottom: var(--space-xs);
+    cursor: pointer;
+    transition: background 0.15s ease;
+}
+
+.goap-agent-item:hover {
+    background: var(--bg-hover);
+}
+
+.goap-agent-item.selected {
+    background: rgba(90, 140, 200, 0.2);
+    border-left: 3px solid var(--accent-sapphire);
+}
+
+.goap-agent-item .agent-name {
+    font-weight: 600;
+    margin-bottom: 2px;
+}
+
+.goap-agent-item .agent-action {
+    font-size: 0.75rem;
+    font-family: var(--font-mono);
+    color: var(--text-secondary);
+}
+
+.goap-agent-item .agent-goal {
+    font-size: 0.7rem;
+    color: var(--accent-sapphire);
+    margin-top: 2px;
+}
+
+.goap-plan-view {
+    flex: 1;
+    overflow-y: auto;
+}
+
+.goap-world-state {
+    display: grid;
+    grid-template-columns: repeat(4, 1fr);
+    gap: var(--space-sm);
+    margin-bottom: var(--space-md);
+}
+
+.goap-stat-card {
+    background: var(--bg-elevated);
+    border-radius: var(--radius-sm);
+    padding: var(--space-sm);
+    text-align: center;
+}
+
+.goap-stat-card .label {
+    font-size: 0.65rem;
+    color: var(--text-muted);
+    text-transform: uppercase;
+}
+
+.goap-stat-card .value {
+    font-size: 1.1rem;
+    font-family: var(--font-mono);
+    font-weight: 600;
+}
+
+.goap-stat-card .bar {
+    height: 3px;
+    background: var(--bg-deep);
+    border-radius: 2px;
+    margin-top: 4px;
+    overflow: hidden;
+}
+
+.goap-stat-card .bar-fill {
+    height: 100%;
+    border-radius: 2px;
+    transition: width 0.3s ease;
+}
+
+.goap-stat-card.thirst .bar-fill { background: var(--stat-thirst); }
+.goap-stat-card.hunger .bar-fill { background: var(--stat-hunger); }
+.goap-stat-card.heat .bar-fill { background: var(--stat-heat); }
+.goap-stat-card.energy .bar-fill { background: var(--stat-energy); }
+
+.goap-plan-steps {
+    background: var(--bg-elevated);
+    border-radius: var(--radius-sm);
+    padding: var(--space-md);
+    margin-bottom: var(--space-md);
+}
+
+.goap-plan-steps h5 {
+    font-size: 0.75rem;
+    color: var(--text-muted);
+    margin-bottom: var(--space-sm);
+}
+
+.goap-plan-flow {
+    display: flex;
+    align-items: center;
+    flex-wrap: wrap;
+    gap: var(--space-sm);
+}
+
+.goap-step-node {
+    padding: var(--space-sm) var(--space-md);
+    background: var(--bg-secondary);
+    border: 2px solid var(--border-color);
+    border-radius: var(--radius-sm);
+    font-family: var(--font-mono);
+    font-size: 0.85rem;
+}
+
+.goap-step-node.current {
+    border-color: var(--accent-emerald);
+    background: rgba(74, 156, 109, 0.15);
+    color: var(--accent-emerald);
+}
+
+.goap-step-arrow {
+    color: var(--text-muted);
+    font-size: 1.2rem;
+}
+
+.goap-goal-result {
+    padding: var(--space-sm) var(--space-md);
+    background: rgba(139, 111, 192, 0.15);
+    border: 2px solid #8b6fc0;
+    border-radius: var(--radius-sm);
+    color: #8b6fc0;
+    font-weight: 600;
+    font-size: 0.85rem;
+}
+
+.goap-inventory {
+    display: grid;
+    grid-template-columns: repeat(3, 1fr);
+    gap: var(--space-xs);
+}
+
+.goap-inv-item {
+    display: flex;
+    align-items: center;
+    gap: var(--space-xs);
+    padding: var(--space-xs) var(--space-sm);
+    background: var(--bg-elevated);
+    border-radius: var(--radius-sm);
+    font-size: 0.75rem;
+}
+
+.goap-inv-item .count {
+    margin-left: auto;
+    font-family: var(--font-mono);
+    font-weight: 500;
+}
+
+.goap-actions-list {
+    flex: 1;
+    overflow-y: auto;
+}
+
+.goap-action-item {
+    display: flex;
+    align-items: center;
+    padding: var(--space-xs) var(--space-sm);
+    border-radius: var(--radius-sm);
+    margin-bottom: 2px;
+    font-size: 0.8rem;
+}
+
+.goap-action-item.valid {
+    background: var(--bg-elevated);
+}
+
+.goap-action-item.invalid {
+    opacity: 0.4;
+}
+
+.goap-action-item.in-plan {
+    background: rgba(74, 156, 109, 0.15);
+    border-left: 3px solid var(--accent-emerald);
+}
+
+.goap-action-item .action-name {
+    flex: 1;
+    font-family: var(--font-mono);
+}
+
+.goap-action-item .action-cost {
+    font-size: 0.7rem;
+    color: var(--text-muted);
+}
+
+.goap-action-item .action-order {
+    width: 18px;
+    height: 18px;
+    background: var(--accent-emerald);
+    border-radius: 50%;
+    display: flex;
+    align-items: center;
+    justify-content: center;
+    font-size: 0.65rem;
+    font-weight: 700;
+    color: var(--bg-deep);
+    margin-right: var(--space-sm);
+}
+
+.no-selection-text, .loading-text {
+    color: var(--text-muted);
+    font-size: 0.85rem;
+    text-align: center;
+    padding: var(--space-lg);
+}
+
+.btn-mini {
+    background: var(--bg-elevated);
+    border: 1px solid var(--border-color);
+    border-radius: var(--radius-sm);
+    color: var(--text-secondary);
+    cursor: pointer;
+    font-family: inherit;
+    transition: all 0.15s ease;
+}
+
+.btn-mini:hover {
+    background: var(--bg-hover);
+    color: var(--text-primary);
+}
+
+.agent-goap-info {
+    margin-top: var(--space-sm);
+    padding: var(--space-sm);
+    background: var(--bg-deep);
+    border-radius: var(--radius-sm);
+    border: 1px solid var(--border-color);
+}
+
+.goap-urgency {
+    display: inline-block;
+    width: 6px;
+    height: 6px;
+    border-radius: 50%;
+    margin-left: 4px;
+}
+
+.goap-urgency.none { background: var(--accent-emerald); }
+.goap-urgency.low { background: var(--accent-gold); }
+.goap-urgency.high { background: var(--accent-ruby); }
+
+@media (max-width: 1400px) {
+    .goap-grid {
+        grid-template-columns: 200px 1fr 250px;
+    }
+}
+
+@media (max-width: 1000px) {
+    .goap-grid {
+        grid-template-columns: 1fr;
+        grid-template-rows: auto;
+    }
+
+    .goap-agents-panel,
+    .goap-goals-panel {
+        grid-row: auto;
+    }
+
+    .goap-panel {
+        max-height: 300px;
+    }
 }