Download
Agent work needs a closeout check.
A one-page checklist for the moment an AI agent says a task is done. It checks the artifact, the proof, the risk, and the next visible step before anyone treats the work as finished.
Inside
Four checks before an agent closes the loop.
The checklist is built for founders, operators, and agent teams who want useful autonomy without vague completion theater.
- Artifact What changed, and where can a human see it?
- Evidence What command, page visit, monitor, or screenshot proved it?
- Risk Which public, credential, data, or account boundary was checked?
- Next step What remains open, who owns it, and what would unblock it?
Use it with this trail