Cognee

Verified

Builds queryable memory for agents by turning documents and conversations into a knowledge graph plus vector store, self-hosted or via API and MCP.

Skills

Memory Ingestion
Ingests documents, code and chat history into a combined graph and vector representation for later recall.
Graph-Aware Search
Answers queries by traversing entity relationships alongside vector similarity for grounded multi-hop recall.
Memory MCP Server
Serves stored memory to Claude and other MCP clients so agents read and write long-term memory as a tool.

Memory & State Vector Databasesagent-memoryknowledge-graphgraph-ragdocument-ingestionself-hostedsemantic-retrieval

Builds a graph-based retrieval index over documents so agents can answer questions spanning many sources, served…

Indexes repositories into a persistent tree-sitter knowledge graph of functions, classes, call chains, and HTTP routes…

Long-term memory store for AI assistants. Extracts and indexes facts, entities, and summaries from conversations. Fast…

Memory API for AI applications. Store and retrieve user memories, bookmarks, and knowledge with semantic search. One…

cognee

Builds queryable memory for agents by turning documents and conversations into a knowledge graph plus vector store, self-hosted or via API and MCP.

fields

nameCognee

providerTopoteretes

urlhttps://github.com/topoteretes/cognee

categoriesmemory-state · vector-databases

accessapi · mcp · cli

authapiKey

streamingfalse

pushfalse

verifiedtrue

officialfalse

tagsagent-memory, knowledge-graph, graph-rag, document-ingestion, self-hosted, semantic-retrieval

skills

memory-ingestionMemory IngestionIngests documents, code and chat history into a combine…

graph-searchGraph-Aware SearchAnswers queries by traversing entity relationships alon…

memory-mcpMemory MCP ServerServes stored memory to Claude and other MCP clients so…