Skip to content

Core Concepts Overview

The Degent Civilization is built around several key concepts that work together to create a dynamic and interactive virtual town environment. This overview will introduce you to these core concepts and how they interact with each other.

System Architecture

flowchart TD
    Start[Start Simulation] --> Init[Initialize Simulation]
    Init --> LoadConfig[Load Config Files]
    LoadConfig --> LoadLLM[Load LLM Configuration]
    LoadLLM --> LoadData[Load Frontend/Backend Data]

    LoadData --> CreateManagers[Create Character State Managers]
    CreateManagers --> UpdateLoop[Update Loop]

    UpdateLoop --> HandleMsg{Handle Server Messages}
    HandleMsg -->|New Day| ResetCountdown[Reset Newday Countdown]
    HandleMsg -->|Other Messages| FilterMsg[Filter Messages by State]

    ResetCountdown --> UpdateStates
    FilterMsg --> UpdateStates[Update Character States]

    UpdateStates --> StateChains[Execute State Chains]

    StateChains --> EnterChain[Enter State Chain]
    StateChains --> ExitChain[Exit State Chain]
    StateChains --> UpdateChain[Update State Chain]
    StateChains --> PostLLMChain[Post LLM Chain]

    EnterChain --> |Functions|BuildPrompt[Build Prompt]
    BuildPrompt --> CallLLM[Call LLM]

    CallLLM --> ProcessResponse[Process LLM Response]
    ProcessResponse --> StateRouter{State Router}

    StateRouter -->|Change State| NextState[Change to Next State]
    StateRouter -->|Stay| UpdateLoop

    NextState --> UpdateLoop

    UpdateLoop --> |Continue|HandleMsg
    UpdateLoop --> |End|End[End Simulation]

    subgraph "State Functions"
        EnterChain
        ExitChain
        UpdateChain
        PostLLMChain
    end

    subgraph "LLM Processing"
        BuildPrompt
        CallLLM
        ProcessResponse
    end

Key Components

1. Simulation

The Simulation is the main container and orchestrator of the simulation. It: - Manages all characters and buildings - Coordinates events and interactions - Maintains the simulation state - Handles time progression - Manages global resources

2. Characters

Characters are the intelligent agents within the simulation. They: - Have personalities and traits - Maintain relationships with other characters - Follow daily routines and schedules - Interact with other characters and buildings - Store and retrieve memories - Change states based on activities

3. Buildings

Buildings are locations within the simulation where: - Characters can be located - Activities take place - Events can be hosted - Resources can be stored - Services can be provided

4. State Management

The state system: - Controls character behavior - Manages transitions between activities - Handles interruptions and priorities - Ensures realistic behavior patterns - Coordinates multi-character interactions

5. Memory System

The memory system allows characters to: - Store experiences and information - Retrieve relevant memories - Influence decision making - Build relationships over time - Learn from past interactions

Component Interactions

Character-Building Interaction

sequenceDiagram
    participant C as Character
    participant B as Building
    participant S as Simulation

    C->>S: Request location change
    S->>B: Check availability
    B->>S: Confirm space
    S->>C: Update location
    C->>B: Perform activity

Character-Character Interaction

sequenceDiagram
    participant C1 as Character 1
    participant SM as State Manager
    participant C2 as Character 2

    C1->>SM: Request interaction
    SM->>C2: Check availability
    C2->>SM: Confirm state
    SM->>C1: Enable interaction
    C1->>C2: Interact

Memory-State Interaction

sequenceDiagram
    participant C as Character
    participant M as Memory System
    participant S as State Manager

    C->>S: Current situation
    S->>M: Request relevant memories
    M->>S: Provide context
    S->>C: Determine action

System Flow

  1. Initialization
  2. Town creation
  3. Character and building setup
  4. State system initialization

  5. Memory system preparation

  6. Runtime

  7. Time progression
  8. State updates
  9. Interaction processing
  10. Memory management

  11. Event handling

  12. Interaction Processing

  13. State checking
  14. Memory retrieval
  15. Action determination
  16. Response generation
  17. Memory storage

Integration Points

1. LLM Integration

The system integrates with Large Language Models for: - Natural language processing - Decision making - Response generation - Memory summarization - Personality expression

2. Database Integration

Persistent storage is used for: - Character data - Memory vectors - Relationship graphs - State history - Event logs

3. External Services

The system can connect with: - WebSocket for real-time updates - Redis for caching - Milvus for vector search - Custom APIs for extended functionality

Next Steps