Pressure-test whether a local fix reveals a broader refactor opportunity. Start skeptical: recommend broader refactoring only when evidence shows it leaves the design net-simpler.
If the target is ambiguous, infer it; ask only if multiple plausible targets would materially change the recommendation.
Workflow
- Summarize the related issue or fix.
- Separate confirmed facts from assumptions.
-
Classify the design signal.
- Choose the best fit:
- simple local mistake
- weak validation or invariant gap
- unclear ownership or boundary problem
- weak type or schema model
- state, lifecycle, concurrency, or ordering hazard
- over-abstraction, under-abstraction, or duplicated logic
- brittle integration or contract mismatch
- If the issue is only a local mistake, say so.
- Choose the best fit:
-
Decide whether a broader refactor is justified.
- Prefer one best recommendation. Include a second only if clearly distinct and worth considering now.
- Good recommendations may delete, inline, collapse, rename, move ownership, or tighten a type/schema.
- Do not equate broader refactoring with adding helpers, adapters, services, interfaces, frameworks, compatibility layers, speculative extension points, or broad rewrites.
- Pick one verdict:
Recommended now;Defer(justified but not worth doing now);Decision needed(justified but hinges on a user tradeoff);No broader refactor recommended(say so and explain why).
-
Scope the recommendation when a broader refactor is justified.
- Explain what boundary, invariant, or ownership concern improves.
- Name the affected code and what happens to it.
- Explain why it beats keeping only the local fix.
- Do not apply code changes unless the user explicitly asks for implementation. Mention tiny obvious cleanups separately if relevant.
Output
## Related issue or fix
## Design signal
## Refactor recommendation
Recommended now | Defer | Decision needed | No broader refactor recommended
### Recommendation, if a broader refactor is justified
### Why this is not over-abstraction, if recommending a refactor
### Scope and affected files, if recommending a refactor
### Benefit, risk, and effort, if recommending a refactor
### Validation, if recommending a refactor
## Alternative considered, if any
## Recommended next step