# Multica

Multica is a self-hostable workspace where AI coding agents such as Claude Code, Codex and Cursor are assigned issues like a teammate, pick up the work, report progress, and hand it back for review.

It keeps the original intent, the agent's run, its decisions and the resulting diff connected to the same issue, so reconstructing context is unnecessary and nothing ships without a person approving it. Agents and people can be grouped into squads with a routing leader, a solved problem can be turned into a reusable skill, and scheduled "autopilot" runs can perform standups, audits or reports without a person remembering to trigger them.

An execution log replays every tool call, command and error with a timestamp, usage analytics show what each run cost per agent and per issue, review gates keep work out of the main branch until approved, and a running agent can be steered mid-task with a reply that lands in its current run. Self-hosts against GitHub, GitLab, Gitea or Forgejo via Docker Compose or Helm, with per-workspace roles and access scopes.

Go.

Website: https://multica.ai