Que fait Add graph-backed memory and context retrieval to agent workflows ?
Use Cognee to ingest project knowledge into graph and vector memory so agents can retrieve durable context across sessions and workflows.
Prerequisites
Python, Cognee, LLM provider credentials, optional graph/vector database backend
Installation
Prerequisite: Python 3.10 to 3.14.
Install Cognee with uv or pip:
- uv pip install cognee
- pip install cognee
For Claude Code persistent memory, upstream documents installing the Cognee integration plugin:
-
claude plugin marketplace add topoteretes/cognee-integrations
-
claude plugin install cognee-memory@cognee
-
Extracted from upstream docs: https://raw.githubusercontent.com/topoteretes/cognee/HEAD/README.md