How surypratapsingh872/chronogenesis-ai works
ChronoGenesis-AI is an experimental platform for causal world modeling, multi-agent analysis, and exploring possible futures. Its principal workflow takes a task through language processing and knowledge retrieval, builds graph and causal context, delegates work to specialist agents, and produces decision, scenario, risk, and report outputs. The README documents those product capabilities; sampled manager and research code confirms several orchestration and analysis calls. Other branches are represented at subsystem level where source is listed but implementation was not sampled, so their internal wiring is not asserted.
Main entry points: manager_agent.py Manager Agent · research_agent.py Research Agent · graph_builder.py Knowledge Graph · retriever.py RAG Retrieval
Components 24
Product Interface
- HTTP API
backend/api/v1 - WebSocket Updates
backend/websocket/routes.py
Agent Orchestration
- Manager Agent
backend/agents/manager_agent.py - Task NLP
backend/services/nlp/nlp_pipeline.py - Specialist Agents
backend/agents - Research Agent
backend/agents/research_agent.py - Decision Agent
backend/agents/decision_agent.py
Knowledge and Reasoning
- Knowledge Graph
backend/services/graph/graph_builder.py - Causal Reasoning
backend/services/causal_inference/causal_reasoning.py - RAG Retrieval
backend/services/rag/retriever.py - Memory Services
backend/services/memory - Transformer Tools
backend/services/transformers - World Model
backend/services/world_model/world_model_engine.py - Environment Model
backend/services/world_model/environment_model.py - Vision Analysis
backend/services/vision/vision_engine.py
Future Exploration
- Counterfactual Engine
backend/services/counterfactual/counterfactual_engine.py - Scenario Simulation
backend/services/simulation - Digital Twin
backend/services/digital_twin/digital_twin_engine.py - Future Prediction
backend/services/deep_learning/success_predictor.py
Platform Services
- Report Generation
backend/services/reporting/report_engine.py - Postgres
backend/database/postgres.py - Neo4j Graph Store
backend/database/neo4j.py - Chroma Store
backend/database/chroma.py
Other components
- User
Connections 29
- User → HTTP API: submits task
- HTTP API → Manager Agent: dispatches task
- User → WebSocket Updates: sends updates
- Manager Agent → Task NLP: processes task
backend/agents/manager_agent.py - Manager Agent → Memory Services: retrieves context
backend/agents/manager_agent.py - Manager Agent → RAG Retrieval: retrieves context
backend/agents/manager_agent.py - Manager Agent → Knowledge Graph: builds graph
backend/agents/manager_agent.py - Manager Agent → Causal Reasoning: analyzes causes
backend/agents/manager_agent.py - Manager Agent → Specialist Agents: delegates analysis
backend/agents/manager_agent.py - Manager Agent → Decision Agent: requests decision
backend/agents/manager_agent.py - Research Agent → Memory Services: retrieves context
backend/agents/research_agent.py - Research Agent → Knowledge Graph: builds graph
backend/agents/research_agent.py - Research Agent → RAG Retrieval: retrieves evidence
backend/agents/research_agent.py - Research Agent → Transformer Tools: reranks results
backend/agents/research_agent.py - Research Agent → Causal Reasoning: analyzes causes
backend/agents/research_agent.py - Research Agent → Environment Model: analyzes setting
backend/agents/research_agent.py - Research Agent → Future Prediction: predicts future
backend/agents/research_agent.py - Knowledge Graph → Causal Reasoning: provides graph
- Knowledge Graph → Neo4j Graph Store: persists graph
- RAG Retrieval → Chroma Store: searches vectors
- Manager Agent → Counterfactual Engine: explores alternatives
backend/agents/manager_agent.py - Manager Agent → Scenario Simulation: simulates scenarios
- Manager Agent → Report Generation: generates report
- Manager Agent → Postgres: stores product data
- User → Vision Analysis: provides visual input
- World Model → Environment Model: uses environment
backend/services/world_model/world_model_engine.py - Manager Agent → World Model: models world
backend/agents/manager_agent.py - Manager Agent → Digital Twin: models twin
backend/agents/manager_agent.py - User → Report Generation: receives results