cognee and subcog

These tools are **complementary**: Cognee is a knowledge engine for AI agent memory, focusing on the core knowledge representation and retrieval, while subcog is a persistent memory system specifically designed for AI coding assistants, suggesting it could integrate with and leverage Cognee for its deeper knowledge processing requirements while providing specialized features like capturing coding session context and hybrid search.

cognee
80
Verified
subcog
43
Emerging
Maintenance 25/25
Adoption 10/25
Maturity 25/25
Community 20/25
Maintenance 13/25
Adoption 6/25
Maturity 9/25
Community 15/25
Stars: 13,204
Forks: 1,336
Downloads:
Commits (30d): 372
Language: Python
License: Apache-2.0
Stars: 17
Forks: 4
Downloads:
Commits (30d): 0
Language: Rust
License: MIT
No risk flags
No Package No Dependents

About cognee

topoteretes/cognee

Knowledge Engine for AI Agent Memory in 6 lines of code

Combines vector search with graph databases to index documents by semantic meaning and learned entity relationships, enabling hybrid retrieval that improves context relevance for agents. Supports multimodal ingestion across arbitrary data formats and structures while maintaining local execution, ontology grounding, and audit trails for trustworthy agent isolation. Integrates with multiple LLM providers and includes CLI tooling and web UI for pipeline management alongside programmatic Python APIs.

About subcog

zircote/subcog

Persistent memory system for AI coding assistants. Captures decisions, learnings, and context from coding sessions. Features hybrid search (semantic + BM25), MCP server integration, SQLite persistence with knowledge graph, and proactive memory surfacing. Written in Rust.

Implements three-layer storage (SQLite + FTS5 + HNSW vectors) with Reciprocal Rank Fusion for hybrid search scoring, exposing ~22 MCP tools via stdio or HTTP transport for Claude Desktop and other AI agents. Achieves 97% factual recall accuracy on LongMemEval benchmarks with automatic all-MiniLM-L6-v2 embedding generation, faceted memory organization by project/branch/file, and optional entity extraction with knowledge graph inference.

Scores updated daily from GitHub, PyPI, and npm data. How scores work