cognithor

Cognithor · Agent OS: Local-first autonomous agent operating system. 19 LLM providers, 18 channels, 145 MCP tools, 6-tier memory, Agent Packs marketplace, zero

What is cognithor?

Cognithor · Agent OS: Local-first autonomous agent operating system. 19 LLM providers, 18 channels, 145 MCP tools, 6-tier memory, Agent Packs marketplace, zero telemetry. Python 3.12+, Apache 2.0.

Overview of cognithor

What it is

Cognithor is a local-first, autonomous agent operating system for AI experimentation and personal automation. It runs entirely on the user's machine, with Ollama or LM Studio, and does not require API keys. Cloud providers are optional. The system is under active development as a pre-v1.0 Beta, with 17,000+ test functions and an 89% coverage gate, but has not been battle-tested in production environments. It is described as a local, private, self-improving assistant that can plan multi-step tasks, evaluate its own results, learn from mistakes, and improve autonomously.

Key features

  • Local-first: runs entirely on the user's machine; works fully offline with Ollama or LM Studio.
  • Supports 19 LLM providers, 18 channels (including terminal, web UI, Telegram, Discord), and 145 MCP tools across 30 modules.
  • 6-tier memory system with 4-channel hybrid search, knowledge vault, and encryption at rest using SQLCipher (AES-256).
  • Agent Packs marketplace for community-contributed skills, with a TUF-Light signed pack registry.
  • Computer Use: desktop automation via screenshots, clicks, keystrokes, and window automation.
  • Zero telemetry: no data sent to the cloud unless a cloud LLM provider is explicitly configured.
  • Audit & compliance features: HMAC + Ed25519 signatures, RFC 3161 TSA, GDPR Art. 15/33, WORM-ready, hash-chained audit trail.

Use cases

  • Personal automation: automating repetitive tasks on the desktop, such as clicking, typing, and reading windows.
  • AI experimentation: testing and comparing multiple LLM providers locally without sending data to the cloud.
  • Privacy-focused AI assistant: keeping all conversations, memories, documents, and credentials on the user's hardware.
  • Multi-step task planning: decomposing complex tasks, evaluating results, and learning from mistakes autonomously.
  • Cross-platform social listening: scanning Reddit, Hacker News, and Discord for leads, with LLM-scored results.

Getting started

See the repository README for installation instructions.

License and pricing

Licensed under Apache-2.0. Open source and free to use.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/alex8791-cybercognithor.svg)](https://yunodb.com/tools/alex8791-cybercognithor?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/alex8791-cybercognithor?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/alex8791-cybercognithor.svg" alt="Featured on YunoDB" height="20"></a>