By topoteretes
Persist conversation context across Claude Code sessions using a knowledge graph, enabling automatic recall of relevant memories and filtered search by user, project, or agent categories.
Store data permanently in the Cognee knowledge graph. Accepts a data category (user, project, or agent) to tag the data with the correct node_set for filtered retrieval.
Search Cognee memory. Session memory is automatically searched on every prompt via hooks. Use this skill explicitly for permanent knowledge graph search, filtered category search, or when you need more results than the automatic lookup provides.
Sync session cache entries into the permanent Cognee knowledge graph. Run this to make session memory searchable, or it runs automatically at session end.
Executes bash commands
Hook triggers when Bash tool is used
Modifies files
Hook triggers on file write and edit operations
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Uses power tools
Uses Bash, Write, or Edit tools
Uses power tools
Uses Bash, Write, or Edit tools
Monorepo for all Cognee-owned integration packages.
Each integration lives under integrations/<name>/ and is an independently publishable package.
integrations/
openclaw/ -> @openclaw/memory-cognee (npm)
claude-code/ -> Cognee plugin for Claude Code
codex/ -> Cognee plugin marketplace for Codex
(Template coming soon. For now, follow the TypeScript pattern below and adapt for Python with pyproject.toml.)
integrations/<name>/ with package.json, entry file, and plugin manifestintegrations/inventory.ymlCI auto-detects new integrations by language (Python via pyproject.toml, TypeScript via package.json) — no workflow edits needed.
Each integration is developed independently with its own toolchain:
# Python integrations
cd integrations/<name>
uv sync --dev
uv run pytest tests/ -v
uv run ruff check .
# TypeScript integrations
cd integrations/<name>
npm install
npx tsc --noEmit
Python integrations must pin the cognee dependency with a bounded range (e.g., cognee>=0.5.1,<0.6.0). This is enforced by CI via scripts/check_version_pins.py. TypeScript integrations that talk to Cognee via HTTP API are exempt from package pinning but should document compatible Cognee server versions.
When a new cognee version is released:
Each integration is published independently via tag-per-package:
# TypeScript: publishes to npm
git tag openclaw-v2026.2.4 && git push --tags
# Python (when added): publishes to PyPI
# git tag <name>-v<version> && git push --tags
The publish.yml workflow parses the tag, runs tests, and publishes to the appropriate registry.
cognee dependencies in Python integrationsintegrations/inventory.yml tracks all known integrations with ownership, migration status, package names, and version info. Update it when adding or migrating integrations.
npx claudepluginhub topoteretes/cognee-integrations --plugin cognee-memoryPersonal knowledge graph for Claude Code — remembers decisions, searches past work, captures sessions
Memory compression system for Claude Code - persist context across sessions
The bridge between Claude's working memory and Basic Memory's durable knowledge graph — session briefings, pre-compaction checkpoints, and capture reflexes
Persistent memory across Claude Code sessions using Cognis
Automatic CORE memory search and ingest plugin for Claude Code
Persistent memory across Claude Code sessions