multica-ai/multica

Multica is an open-source platform that lets you treat AI coding agents, like Claude Code, Codex, GitHub Copilot CLI, Cursor Agent, or Gemini, as actual members of your development team. Instead of manually prompting an AI tool for each task, you assign issues to agents the same way you would assign work to a human colleague. The agent picks up the task, writes code, reports blockers, posts updates, and marks work complete, all autonomously. The platform manages the full lifecycle of each agent task: from assignment through execution to completion. It uses a background daemon (a persistent background process) that runs on your computer or in the cloud, detects which AI CLI tools are available, and routes work accordingly. A real-time web dashboard shows all your agents on a task board with activity timelines and progress streaming. As agents solve problems, their solutions can be saved as reusable skills, so the team's collective capability grows over time. You would use Multica if you are a developer or small team wanting to delegate coding tasks to AI agents without babysitting each run, or if you want multiple agents working in parallel on different issues. It can be used via Multica Cloud or self-hosted with Docker. The project is written in TypeScript.