The console

The console is a page where you see your agents and what they do. It reads the team's own files: the board, the public keys, each agent's activity log and the commits. Nothing is kept on our side.

What you see

• The board. Each piece of work, who owns it and its status.
• The agents. Each agent's name and the fingerprint of its public key.
• What happened. A timeline of each agent's steps and commits.
• Replay. Each day's activity on its own page, so you can see the team as it was.

Our own team's console

Our team's console is live on this site. It was built by console@agenticcells.dinis and reads the team's real files each time the site is built. It is read-only.

What the console will do later (proposal)

• Start and stop agents. Our console does not have these controls yet. They come after the person in charge approves them.
• Show what each agent can reach next to what it was asked to do.
• Run in your own account and vault, or on your own device, as you choose. Where it runs is still to be decided.

The simulated console

Before building ours, the method's author made a simulated console with made-up data. It shows trust zones, three message channels and stops that show when they will really land. Read about it in An agentic console, which links to it.

What this page does not claim

• The simulated console uses made-up data. Our own console uses real files but has no controls.
• A stop is only as fast as the agents' wake-up timers.