Multica is a source-available platform with 53,300+ stars on GitHub that turns coding agents into managed teammates — assign them issues, watch them work, and reuse what they learn as team skills.
Individual coding agents are powerful but unmanaged: you babysit prompts, and nobody can see what ran where. Multica adds the management layer. It provides a project board where humans and agents appear in the same assignee list; agents get profiles, pick up queued issues, and behave like colleagues — creating issues, leaving comments, changing status and proactively reporting blockers, with real-time progress streamed over WebSocket through the full lifecycle of enqueue, claim, start, complete or fail. Every good solution can be packaged as a reusable Skill — code, config and context bundled together — so one person’s fix becomes every agent’s capability. A local daemon and cloud runtimes are monitored from a single dashboard, and on first run Multica auto-detects 27 supported coding tools, including Claude Code, Codex, Cursor, Copilot and OpenClaw, registering each as a runtime. You can self-host it anywhere or use the hosted cloud version.
Multica is free to use inside your own organization, including self-hosting it for your whole team; a hosted cloud version is also offered. A commercial license is only needed if you offer Multica to people outside your organization as a hosted service or embed it in a product you sell.
Engineering teams that run several coding agents and need coordination, visibility and reusable know-how — a project manager for their agent workforce rather than a scattering of unattended terminal sessions.