Command 'em. Then check 'em.
Chat with Claude, Codex, Gemini and more across every project. Commandem shows what needs you, runs the checks you set, and keeps the receipts.

Example data. Screenshots, not a live app.
Chat
Every agent, one conversation.
Talk to Claude, Codex, Gemini and other agents from one window, with each project's history kept in place. When an agent asks to run a command, the request can wait for your Allow or Deny.

Plan
Name the check before the work starts.
Each Mission holds the goal, its tasks, and the check that decides when it is done. The board shows what is now, next and finished across your projects.

Review
See what changed, and how you know.
Review lists what each agent changed and whether its check passed, in plain words. Work is called verified only when its check failed on the old code and passes on the new.

Works with the agents you already use.
- Claude
- Codex
- Gemini
- Antigravity
- Local models
- ACP agents
- Grok (experimental)
Support differs by agent. See what each one supports.
Runs on your Mac.
Your projects, history and records stay on hardware you control.
- No Commandem accountUse your existing agent sign-ins. There is nothing to register with us.
- No telemetryCommandem does not report usage or send analytics anywhere.
- No hosted relayModel providers receive only what you send them.
Public beta coming November 26
Free and open source, for macOS 14 or later on Apple silicon. Join the Discord to hear the moment it ships.