setup-cognee

Configure and verify cognee MCP persistent memory for Claude Code.

2|Updated Feb 23, 2026
One-click install
npx skills add https://github.com/MDGrey33/memnyx --skill setup-cognee
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: setup-cognee
Source: https://github.com/MDGrey33/memnyx/tree/main/.claude/skills/setup-cognee
Command: npx skills add https://github.com/MDGrey33/memnyx --skill setup-cognee

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill removes the complexity of installing, configuring, and validating cognee-based persistent memory for Claude Code so the agent can remember context across sessions without manual setup drift.

Core Features & Use Cases

  • Interactive setup: Detects the local environment, available tools, and existing config before guiding installation.
  • Memory backend configuration: Supports file-based or Postgres-backed storage, plus LLM and embedding provider selection.
  • Safe secret handling: Uses macOS Keychain or a literal fallback path depending on platform, while keeping generated config aligned.
  • Lifecycle management: Handles initial setup, refreshes after updates, and post-restart health checks.
  • Use case: A developer can initialize a new Claude Code workspace, enable semantic memory, and confirm the system works with a real remember-and-recall probe.

Quick Start

Ask the skill to set up cognee MCP for your machine, choose the storage and model options, and verify the installation with a successful remember-and-recall round trip.

Frequently Asked Questions about setup-cognee

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I set up persistent memory for Claude Code across sessions?

To set up persistent memory for Claude Code, you configure the cognee MCP server by editing the canonical .mcp.json file, aligning the generated .env file, and securely handling your API keys. This enables the agent to remember context across sessions without manual setup drift.

Does Claude Code MCP setup support storing memory in Postgres?

Yes, Claude Code MCP setup supports Postgres-backed storage for persistent memory. During configuration, you can select between file-based storage or a Postgres backend, along with your preferred LLM and embedding provider for semantic retrieval.

What is the best way to configure cognee MCP API keys on macOS?

The best way to configure cognee MCP API keys on macOS is by storing them securely in the Keychain. The setup process detects your local environment and uses the macOS Keychain for safe secret handling, applying a literal fallback path for non-macOS systems.

How do I verify my Claude Code semantic memory installation is working?

You verify your Claude Code semantic memory installation by running an end-to-end stdio remember and recall probe. This lifecycle health check confirms that the cognee MCP configuration is correctly applied and functioning after setup or system restarts.

Why does my Claude Code agent forget context after restarting?

Your Claude Code agent forgets context after restarting because persistent memory has not been configured or validated. You need to initialize the cognee MCP backend, select your storage and model providers, and run a post-restart health check to repair setup drift.

Can I refresh an existing cognee MCP configuration after updating Claude Code?

Yes, you can refresh an existing cognee MCP configuration after updates. The setup process handles lifecycle management by applying drift repair, regenerating the .env alignment, and running post-restart health checks to ensure continuous semantic memory operation.