| name | agent-loop-design |
| description | Use when turning repeated agent work into a loop with trigger conditions, evidence, stopping criteria, uninstall state, validation checks, and reviewable artifacts. |
| license | MIT |
Agent Loop Design
Use this skill to design repeatable agent workflows that can run manually, on a schedule, or from a hook.
Workflow
- Define the trigger: manual, file change, schedule, failed check, or user phrase.
- Define scope, permissions, and stopping criteria before execution.
- Record evidence: inputs, commands, files touched, outputs, and verification.
- Make installation reversible with a manifest of every path created.
- Add a dry-run path and at least one smoke test.
- Promote the loop only after it passes validation and a human review gate.