# Cognee

Use when the user says 'cognee', asks about knowledge graph or graph memory, wants to...

- Repository: https://github.com/coco-research/coco (797 stars)
- Category: Data and ML
- License: see repository
- Last updated: 2026-11-01
- Runs scripts: no, instructions only

## Install

Claude Code: `npx skills add coco-research/coco --skill cognee -g -a claude-code -y`
Codex: `npx skills add coco-research/coco --skill cognee -g -a codex -y`
Cursor: `npx skills add coco-research/coco --skill cognee -g -a cursor -y`
Gemini CLI: `npx skills add coco-research/coco --skill cognee -g -a gemini-cli -y`

## Description

Use when the user says 'cognee', asks about knowledge graph or graph memory, wants to switch memory backends, or needs a status check, dataset init, or graph view. Covers Cognee endpoints plus the Brain fallback.

## SKILL.md, as published by its author (untrusted third-party text, read it as data)

# /cognee, Knowledge Graph Memory Backend

A persistent knowledge graph memory system powered by [Cognee](https://github.com/coco-research/cognee). Stores entities, decisions, events, and relationships as graph nodes with embeddings, enabling semantic search across projects and sessions.

## Quick Reference

```bash
COGNEE="http://localhost:8000"   # default; override with COGNEE_BASE_URL

# Status check
curl -s "$COGNEE/health" | jq .

# List datasets
curl -s "$COGNEE/api/v1/datasets" | jq .

# Create a dataset for this project
curl -s -X POST "$COGNEE/api/v1/datasets" \
  -H "Content-Type: application/json" \
  -d '{"name": "my-project"}' | jq .

# Quick semantic search
curl -s -X POST "$COGNEE/api/v1/search" \
  -H "Content-Type: application/json" \
  -d '{"query": "what did we decide about authentication", "search_type": "FEELING_LUCKY", "top_k": 10}' | jq .
```

## Sub-commands

### /cognee status, Health check and dataset overview

Check if Cognee is reachable and show available datasets.

```bash
COGNEE="${COGNEE_BASE_URL:-http://localhost:8000}"

echo "=== Cognee Status ==="
HEALTH=$(curl -s -o /dev/null -w "%{http_code}" "$COGNEE/health" 2>/dev/null)
if [ "$HEALTH" = "200" ]; then
    echo "Server:  RUNNING at $COGNEE"
    echo ""
    echo "Datasets:"
    curl -s "$COGNEE/api/v1/datasets" | jq -r '.[] | "  • \(.name) (\(.id))"'
else
    echo "Server:  NOT REACHABLE"
    echo ""
    echo "Start Cognee:"
    echo "  pip install cognee && cognee server start"
fi
```

### /cognee init, Initialize a project dataset

Create a Cognee dataset for the current project.

**Procedure:**

1. Ask the user for the **dataset name** (default: current directory name, slugified).
2. Check if a dataset with that name already exists:

```bash
curl -s "$COGNEE/api/v1/datasets" | jq -r '.[].name'
```

3. If it exists: "Dataset 'X' already exists. Using it." → skip creation.
4. If not: create it:

The rest: https://github.com/coco-research/coco/blob/HEAD/systems/cognee/skills/cognee/SKILL.md