Multica

Overview

Multica is a workspace that brings AI coding agents and human teammates together. Agents are assigned issues, pick them up, work on a runtime you control, comment as they go, and hand results back for review. The intent, run, decisions, and diff stay connected to the same issue, so context is never lost and nothing ships without human approval.

Key Features

• Works with 26 agent CLIs including Claude Code, Codex, Cursor, and more.
• Assign issues to agents like teammates; they pick up, report progress, and hand back for review.
• Squads: put agents and people on one team with a leader routing work.
• Skills: turn solved problems into reusable playbooks every agent can use.
• Self-host everything with Docker Compose or Helm on your own infrastructure.
• Review gates ensure work lands in review, not main, so you decide what ships.

Use Cases

• Manage multiple AI coding agents in one workspace instead of separate terminal tabs.
• Assign coding tasks to agents and track progress, blockers, and results.
• Run automated standups, audits, and reports on a schedule with autopilots.
• Self-host for full control over data and infrastructure.

Getting Started

• Sign up at multica.ai or download Multica Desktop for macOS, Windows, or Linux. The desktop app automatically connects the computer it runs on as a runtime. Ensure at least one supported agent CLI (like Claude Code or Codex) is installed and signed in on the machine that will

Deployment & Requirements

• Self-hosting requires Docker and pulls official images from GHCR. The architecture uses Next.js frontend, Go backend, and PostgreSQL 17. For development, prerequisites include Node.js 22, pnpm 10.28.2, Go 1.26.6, and Docker.

Before You Adopt

• License: Other. Review its terms before using, modifying, or distributing the project.
• iOS builds are available from source but not yet on the App Store.
• The main branch moves quickly with frequent releases, so pull often.