| name | self-improving-agent-loop |
| description | Design governed self-improving agent loops that sense work, model failure modes, act in small slices, gate evidence, and turn learnings into tracked proposals. |
Self-Improving Agent Loop
Use this skill when a project wants agents to improve a workflow, skill, prompt, loadout, checker, or operating policy over time.
Loop
- Sense: collect issues, diffs, tests, traces, artifacts, user feedback, and handoff notes.
- Model: identify the task type, capability assumption, failure mode, and quality bar.
- Plan: choose one reversible improvement with explicit evidence.
- Act: change the smallest useful surface using existing project patterns.
- Gate: run tests, artifact checks, permission checks, and review gates.
- Learn: record a follow-up issue, durable memory, skill update, or checker proposal only when evidence supports it.
Governance
- Treat frontier models as capable of planning, synthesis, review, and context work when the digital surfaces are available.
- Keep deterministic scaffolding for repeatable validation, policy checks, manifests, and install wiring.
- Treat generated traces and self-assessments as evidence candidates, not proof.
- Require human approval for permission expansion, destructive actions, production data, publishing, merge, deployment, and policy changes.
- Keep domain-specific loops out of the core pack unless the pattern generalizes across projects.
Output
Loop Goal
Sensors
Failure Model
Reversible Action
Gates
Human Approval Boundaries
Learning Output
Next Issue Or Memory
