I've been building Praxi-Bot, an Apache-2.0-licensed, local-first AI agent platform in Rust and Tauri. I'm opening up development and looking for people who want to build it together.

What is Praxi-Bot?

Praxi-Bot is designed to run on your own machine or server and to connect local or cloud language models with practical automation tools.

  • Rust backend: asynchronous daemon and CLI, with Tokio and Axum.
  • Desktop application: Tauri and React, with an initial Windows alpha installer.
  • Model support: cloud LLM providers and local models through Ollama.
  • Automation: desktop and browser tools, workflows, and scheduled tasks.
  • Integrations: MCP tooling and external services.
  • Permission boundaries: approval checks and local-first data handling.

It's still alpha software. The APIs and implementation details may evolve, and feedback on architecture and reliability is welcome.

Where you can help

I've split several tasks into focused, beginner-friendly GitHub issues with acceptance criteria.

More experienced developers are also welcome to explore MCP compatibility, agent evaluations, Linux and macOS support, and Rust architecture.

How to make your first contribution

  1. Read CONTRIBUTING.md.
  2. Open the curated good first issues or good first issue label.
  3. Leave a comment on the issue you'd like to pick up, so we can coordinate.
  4. Open a small, focused PR, with tests or documentation as appropriate.

AI-assisted contributions are welcome, provided contributors review, understand, and verify the proposed changes. You don't need to be an expert to get started: clear docs, tests, accessibility, and small fixes all matter.

I'm happy to explain the relevant parts of the codebase, discuss technical decisions, and review PRs.

Repository: https://github.com/emreocell/Praxi-Bot

For other maintainers: what made your first outside contributors decide to get involved?