compozy
An operating system for AI agents. The autonomous control plane for your agents: see and steer every Claude Code, Codex, and Cursor session. Always-on agents ch
What is compozy?
An operating system for AI agents. The autonomous control plane for your agents: see and steer every Claude Code, Codex, and Cursor session. Always-on agents check in on their own, and you step in only when one needs you.
Overview of compozy
What it is
CompozyOS is an operating system for AI agents. It functions as an autonomous control plane that lets users see and steer every Claude Code, Codex, and Cursor session. Always-on agents check in on their own, and users step in only when one needs them. It runs the agent CLIs users already have as durable sessions that outlive the terminal, keeps every one of them visible, and calls the user only when a decision is needed. Loops, schedules, permissions, and approvals come built in, so there is no stack to assemble around the agent.
Key features
- Sessions that outlive the terminal: fork a session, continue it with another agent, rewind to a checkpoint, or steer it mid-turn.
- Worktrees give parallel agents their own checkout of the same repository.
- Handoff continues a session with another agent CLI when one runs out.
- Approvals hold a tool call until a person allows it, and waiting work rises to Needs you.
- Always-on agents check in on a schedule, stay quiet when there is nothing to say, and call you when something needs a decision.
- Loops are named contracts with a goal, a way to verify it, and a stop; they carry an iteration cap and can have a human gate.
- Built-in integrations for fifteen agent CLIs including Claude Code, Codex, Cursor Agent, Gemini CLI, Google Antigravity, GitHub Copilot CLI, OpenCode, Pi, Kimi CLI, and Qwen Code, plus model providers and gateways like OpenRouter, Vercel AI Gateway, xAI, z.ai, and Moonshot / Kimi.
Use cases
- Triage new issues as they land: an always-on agent watches new GitHub issues, proposes what to do, and the user picks.
- Deliver a spec while you are away: a conductor splits the work into slices, each slice runs in its own worktree.
- Review a change until the round is clean: a reviewer finds, an implementer fixes, and findings land in a file you can open.
- Leave a briefing waiting every weekday: a schedule starts an agent with nobody watching, every run is recorded with retry on failure.
Getting started
bashcurl -fsSL https://compozy.com/install.sh | sh``````bash# Homebrewbrew install compozy/compozy/compozy# npmnpm install -g @compozy/cli# Go, with the tag shown on https://github.com/compozy/compozy/releases/latestgo install github.com/compozy/compozy@# From sourcegit clone https://github.com/compozy/compozy.gitcd compozygo build -o ./bin/compozy .Then start one session from the repository you want agents to work on:bashcompozy install # pick a provider (the verified installer opens this for you)compozy daemon start # start the local daemoncompozy session new --agent general --name first-run
License and pricing
License: MIT. Pricing type: open source.
Based on the project's public README and repository metadata.