Back to directory

Leon

An open-source personal AI assistant built around tools, context and memory, able to plan and carry out multi-step tasks locally or with remote AI providers rather than only answering with text.

Leon

Leon is an open-source personal AI assistant built around tools, context, memory and agentic execution. It can run in a smart mode that picks its own approach, a controlled mode that follows deterministic native skills and actions, or an agent mode that plans a task step by step.

It supports native skills for controlled actions and SKILL.md-backed agent skills for open-ended workflows, and can use real tools to complete tasks rather than only replying with text. A computer-use capability lets it operate desktop and browser interfaces and verify visual outcomes, running locally or through a companion component. It keeps layered memory covering durable preferences, day-to-day context and recent discussion, and can use both local and remote AI providers.

The project is currently focused on a 2.0 developer preview with agentic execution, while a more stable, pre-agentic release remains available on a separate branch. It is written in TypeScript and Python.

More in Agent frameworks

See all 902
  • Recurse

    recurse.run

    A serverless harness that lets a coding agent design, test, and deploy its own specialist sub-agents as tools, MCP servers, or bots, defined by a schema-validated manifest.

  • ProofForge

    shulgin.is-a.dev

    An AI-agent pipeline that decomposes a mathematics problem, proves each piece, and formalizes the result in Lean 4 and Mathlib until the Lean kernel accepts it, with six of its proofs already merged into Google DeepMind's formal-conjectures repository.

  • local-coder

    github.com

    Builds a local coding-agent team on top of Ollama and OpenCode by detecting the machine's hardware, recommending models for seven specialist roles, and testing whether each assigned model can actually use tools, edit a repository and pass verification.