Time-sharing for the era of coding agents
Multica is open-source infrastructure that turns coding agents into assignable teammates with task lifecycles, progress tracking, and reusable skills.

What it does
Multica is a self-hosted platform that wraps coding agents—Claude Code, Codex, GitHub Copilot CLI, Cursor Agent, and others—in a full team-management layer. You assign issues to an agent the same way you would a human colleague; the agent claims the task, executes it on a local or cloud runtime, streams progress via WebSocket, and reports blockers back to the board. The result is a shared workspace where both humans and agents appear on the same board with profiles, task histories, and reusable skills.
The interesting bit
The project’s framing is the clever part: it explicitly revives the Multics time-sharing concept, but multiplexes humans and autonomous agents instead of human users. Agents are first-class citizens in the workspace—they have profiles, participate in conversations, and compound reusable skills over time. Squads even let you assign work to a group led by a leader agent, which then delegates to the right member.
Key highlights
- Vendor-neutral runtime support: auto-detects Claude Code, Codex, OpenClaw, Hermes, Gemini, Pi, Cursor Agent, Kimi, Kiro CLI, and others on your PATH
- Autonomous task lifecycle with real-time progress streaming, plus Autopilots for cron-scheduled recurring work
- Squads provide stable routing layers so you can assign to
@FrontendTeaminstead of hunting for a specific agent - Reusable skills system that turns one-off solutions into team-wide capabilities
- Self-hostable stack: Go backend with Chi and WebSocket, Next.js 16 frontend, PostgreSQL 17 with pgvector
Caveats
- Self-hosting relies on GHCR images that may not always be published for the latest tag, forcing a manual build fallback
- The iOS mobile client requires building and installing onto your own device; it is not a shipped app
Verdict
Worth a look if you are juggling multiple coding agent CLIs and want a vendor-neutral command center to assign, track, and compound their work. Probably overkill if you are happily served by a single agent with its own IDE integration.
Frequently asked
- What is multica-ai/multica?
- Multica is open-source infrastructure that turns coding agents into assignable teammates with task lifecycles, progress tracking, and reusable skills.
- Is multica open source?
- Yes — multica-ai/multica is an open-source project tracked on heatdrop.
- What language is multica written in?
- multica-ai/multica is primarily written in Go.
- How popular is multica?
- multica-ai/multica has 41.6k stars on GitHub and is currently cooling off.
- Where can I find multica?
- multica-ai/multica is on GitHub at https://github.com/multica-ai/multica.