Skip to content
Get packs

Cognee Integrations Setup

Guides an agent through configuring cognee's LLM, embedding, database, storage, and MCP server settings

Source Download skill
Repository
topoteretes/cognee
Stars
33,002
Licence
Apache-2.0
Files
1 file

What it does

When installed, the agent knows how to connect the cognee memory system to other services: switching the language model or embedding provider (OpenAI, Azure, Gemini, Anthropic, Ollama, OpenRouter, Bedrock), swapping databases (Postgres, pgvector, Neo4j, Neptune, Turso), using S3 storage, and running the MCP server so IDEs like Cursor or Claude Code can use cognee memory. It points the agent to the repo's .env.template as the source of truth and warns about common mistakes, such as configuring the LLM but not the embedding provider, or needing to reset data after changing embedding models.

Great for

  • Developers setting up cognee for a new project who want to use a non-default model or database
  • Teams deploying cognee with Postgres, Neo4j, or S3 and needing correct environment variables
  • Anyone connecting cognee's MCP server to an IDE assistant

Skip it if

  • People who only want to use cognee with its default OpenAI and SQLite/LanceDB setup, since the skill adds little for that case
  • Users who would paste real API keys or cloud credentials into a shared or public chat, since the skill involves handling secrets in .env files
  • Workflows that need guaranteed current variable names without checking the repo, because the skill itself lists some names that may change

Try asking

  • “Switch cognee to use Anthropic for the LLM and keep OpenAI embeddings, and show me the variables to set”
  • “Set up cognee with Postgres for relational data and pgvector for vectors using docker compose”
  • “Connect the cognee MCP server to Claude Code and explain what DB_* settings it needs”
  • “I changed the embedding model in my cognee project; what do I reset and how do I re-ingest?”
Step-by-step guide: how to use Cognee Integrations SetupInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the Apache-2.0 licence and an attribution note.

    cognee-integrations.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip cognee-integrations.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip cognee-integrations.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Check the repo's .env.template before asking the agent to invent variable names
  • When changing the LLM, also set the embedding provider and model so they do not silently fall back to OpenAI
  • Back up data before running forget commands, since they delete graph and vector state

What's inside

  • SKILL.md

Redistributed unmodified from topoteretes/cognee under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.