skills.rest
topoteretes
topoteretesOFFICIAL
Explore all skills by this author →
Website𝕏@cognee_GitHub

cognee-install

Installs cognee and runs a first remember-recall flow with the Python SDK.

30.4k|3.0k|Updated Aug 16, 2023
Search Tags:#AI memory#Installation#lLM#python SDK#Setup#Cognee#Knowledge Graph
One-click install
npx skills add https://github.com/topoteretes/cognee --skill cognee-install
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: cognee-install
Source: https://github.com/topoteretes/cognee/tree/main/.claude/skills/cognee-install
Command: npx skills add https://github.com/topoteretes/cognee --skill cognee-install

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires cognee.

What problem does it solve?

Setting up cognee for the first time involves choosing Python versions, optional extras, LLM provider configuration, and understanding the remember/recall memory API. This Skill walks through a complete fresh installation and a minimal working example so the first memory flow runs correctly.

Core Features & Use Cases

  • Guided Installation: Sets up a virtual environment with uv, installs cognee from PyPI or editable source, and selects extras like postgres, neo4j, docling, anthropic, ollama, or aws only when needed.
  • Configuration Defaults: Explains that only an LLM_API_KEY is required, with SQLite, LanceDB, and Ladybug working locally out of the box.
  • First Memory Flow: Provides a minimal async script using remember() and recall(), plus session memory, dataset scoping, and troubleshooting for caching, feedback, and instructor mode settings.
  • Use Case: A developer wants to give an agent persistent memory; they follow this Skill to install cognee, set their OpenAI key, and run a script that stores a fact and recalls it via natural language query.

Quick Start

Install cognee with uv, set your LLM_API_KEY in a .env file, and run a minimal async script that calls cognee.remember on a sentence and then cognee.recall with a question about it.

Frequently Asked Questions about cognee-install

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

FAQPage Schema
How do I install cognee with Python?▼

Install cognee using uv pip install cognee from PyPI, or uv pip install -e . when working inside the repository. It requires Python 3.10 through 3.14, and a virtual environment created with uv venv is recommended.

How do I run a first cognee remember and recall example?▼

Set LLM_API_KEY in a .env file, then run an async script calling await cognee.remember with your text followed by await cognee.recall with a question. The remember call handles add, cognify, and improve in one step.

What databases does cognee need to run locally?▼

Cognee needs no external services by default. It uses SQLite for relational storage, LanceDB for vectors, and Ladybug for the graph, all stored locally, with OpenAI as the default LLM and embedding provider.

Which cognee extras should I install for postgres or neo4j?▼

Install extras only when needed, such as cognee[postgres], cognee[neo4j], cognee[docling], cognee[anthropic], cognee[ollama], or cognee[aws]. The full list of optional dependencies is in pyproject.toml under project.optional-dependencies.

Why does cognee give structured output errors with my LLM?▼

Structured output errors usually mean the model or provider needs an explicit instructor mode. Set LLM_INSTRUCTOR_MODE to json_schema_mode in your environment to resolve these errors.

How do I reset cognee local state during testing?▼

Run cognee-cli forget --all from the shell, or call await cognee.forget with everything=True in Python. This wipes local state so experiments can start fresh.

Related Skills

Explore More Software Engineering Skills→
thinking-machines-labthinking-machines-lab
4.0k

setup

Guide Tinker installation, environment setup, API key provisioning, and verification.

Official
Intermediate
davila7davila7
30.5k

cocoindex

Build incremental data transformation pipelines for embeddings, knowledge graphs, and search indexes.

Community
Advanced
haxlyshaxlys

vector-index-tuning

Tune HNSW parameters and quantization strategies for Qdrant vector indexes.

Community
Advanced
travelinman1013travelinman1013

chroma

Store and query embeddings with metadata in a local vector database.

Community
Intermediate