How joeydd032995-pixel/self-morphing-adaptive-recursion-engine works
The product is a hybrid neurosymbolic reasoning engine exposed through a CLI and authenticated FastAPI service. A typical user submits a query, which the engine can plan through knowledge-graph hops and verify before returning a result; documents can also be ingested for retrieval-grounded reasoning. The authoritative engine contains morphic execution, RAG, knowledge-graph, self-teaching, and GNN capabilities. Edges reflect only calls or relationships shown in the excerpts or README; unsampled subsystem internals are represented at the engine path.
Main entry points: organized_self_morphing_engine.py Adaptive engine · cli.py CLI · fastapi_server.py FastAPI service · semantic_cache.py Semantic cache
Components 17
User Interfaces
- CLI
cli.py - FastAPI service
fastapi_server.py
Reasoning Engine
- Adaptive engine
organized_self_morphing_engine.py - Morphic execution
organized_self_morphing_engine.py - Hybrid similarity
organized_self_morphing_engine.py - PoG planning
organized_self_morphing_engine.py - Verification and audit
organized_self_morphing_engine.py
Knowledge and Retrieval
- RAG retrieval
organized_self_morphing_engine.py - Document ingestion
organized_self_morphing_engine.py - Knowledge graph
organized_self_morphing_engine.py
Learning and Models
- Self-teaching
organized_self_morphing_engine.py - Learned GNN
organized_self_morphing_engine.py
Persistence
- Semantic cache
semantic_cache.py - SQLite engine store
organized_self_morphing_engine.py
Other components
- User
- Neo4j
- LLM provider
Connections 21
- User → CLI: runs command
cli.py - CLI → Adaptive engine: dispatches to
cli.py - User → FastAPI service: sends request
fastapi_server.py - FastAPI service → Adaptive engine: calls engine
fastapi_server.py - Adaptive engine → Morphic execution: executes nodes
README.md - Adaptive engine → Hybrid similarity: scores queries
README.md - Adaptive engine → PoG planning: plans reasoning
fastapi_server.py - PoG planning → Knowledge graph: explores hops
README.md - PoG planning → Verification and audit: reflects through
organized_self_morphing_engine.py - Adaptive engine → RAG retrieval: retrieves context
fastapi_server.py - Adaptive engine → Document ingestion: ingests documents
cli.py - Document ingestion → Knowledge graph: seeds entities
README.md - Adaptive engine → Knowledge graph: syncs graph
README.md - Knowledge graph → Neo4j: syncs bidirectionally
README.md - Adaptive engine → Self-teaching: runs teaching loop
cli.py - Self-teaching → Verification and audit: validates proposals
README.md - Self-teaching → Learned GNN: uses graph signals
README.md - Adaptive engine → Learned GNN: builds and trains
README.md - Adaptive engine → Semantic cache: gets and puts
organized_self_morphing_engine.py - Adaptive engine → SQLite engine store: persists engine state
organized_self_morphing_engine.py - Adaptive engine → LLM provider: requests completions
README.md