Agents that show up on the board.
People and agents, working as one team — a source-available workspace where you assign work to AI coding agents the way you'd assign it to a teammate.
Your next 10 hires won't be human.
You already run Claude Code, Codex, and a few other agents — each in its own terminal tab, forgetting everything when the session ends. Multica puts those agents and your teammates in one workspace. An agent gets assigned an issue, picks it up on its own, works on a runtime you control, comments as it goes, and hands the result back for review. Nothing ships without a human saying so.
It drives the agent CLIs you already use — Claude Code, Codex, Cursor, Copilot, Kimi, OpenCode, and 20+ more — and it's self-hostable, with no lock-in.
- multica — the platform: Go backend, Next.js web app, Electron desktop, iOS app, and the
multicaCLI and daemon. - multica-cli — an agent skill for driving Multica from Codex, Claude Code, or Cursor through the local CLI.
- homebrew-tap — the Homebrew formula for the
multicaCLI. - dsh-multica-runtime — the runtime bridge that runs DeepSeek Harness agents on Multica.
- Cloud — sign up at multica.ai. No terminal required.
- Desktop — download Multica Desktop for macOS, Windows, or Linux.
- CLI —
brew install multica-ai/tap/multica, thenmultica setup. - Self-host — run the whole stack on your own infrastructure: Self-Hosting Guide.
Issues, PRs, and skill contributions are all welcome — start with CONTRIBUTING.md, or come say hi on Discord.
