Skill · SKILL.md

A Review That Doesn't Rubber-Stamp

A reusable Skill that reviews a change for real problems instead of approving it by reflex.

TypeSKILL.md

You ask the agent to review the change before it ships. It reads the diff and says it looks good — clean, sensible, no concerns. It ships. The bug was in the edge case the diff didn't show: the empty input, the second concurrent call, the one caller still passing the old shape. The review wasn't wrong, exactly. It just never looked where bugs live.

A review that only confirms what the diff shows isn't a review — it's a rubber stamp with extra words. A real one works down a fixed set of dimensions, every time, and is allowed to find nothing only after it says what it checked. Packaging it as a Skill makes that rigor repeatable instead of mood-dependent:

✕ Rubber stamp

Reads the diff top to bottom

▼

"Looks good, no concerns"

▼

The edge case it never opened ships

✓ Real review

Walks correctness → edges → blast radius → readability

▼

Each finding: file, severity, the smallest fix

▼

Nothing slips, because nothing was skipped

Save it as a SKILL.md and invoke it on any change worth a second look. The dimensions matter more than the verdict: correctness you can see, but edges, blast radius, and the next reader are where the costly bugs hide. The one rule it must never break — if it finds nothing, it says what it checked and why it's sure. Approval by default is precisely how the edge case ships.

Save as a Skill

---
name: review-changes
description: Review a diff for real problems before it ships — no rubber-stamping
---
Review the change across these dimensions, in order:
1. **Correctness** — does it do what it claims? Trace one real input through the new path.
2. **Edge cases** — empty, null, huge, the second call, concurrent access. Where does it fall at the edges?
3. **Blast radius** — what else calls this? What silently changes for them?
4. **Readability** — could the next person change it without you in the room?

For each finding, report: `file:line` · severity (blocker / should-fix / nit) · why it matters · the smallest fix.
If you find nothing, say what you checked and why you're confident. Never approve by default.

Prefer your agent install it?

Add this rule to my CLAUDE.md, verbatim:

---
name: review-changes
description: Review a diff for real problems before it ships — no rubber-stamping
---
Review the change across these dimensions, in order:
1. **Correctness** — does it do what it claims? Trace one real input through the new path.
2. **Edge cases** — empty, null, huge, the second call, concurrent access. Where does it fall at the edges?
3. **Blast radius** — what else calls this? What silently changes for them?
4. **Readability** — could the next person change it without you in the room?

For each finding, report: `file:line` · severity (blocker / should-fix / nit) · why it matters · the smallest fix.
If you find nothing, say what you checked and why you're confident. Never approve by default.

Did this help?

Tags#verify #workflow #skill

Other skills

01Verify Before You Call It Done 02Plan First, Then Code 03Map the Blast Radius Before You Edit 04If It's Deterministic, Don't Make It Guess 05Diagnose From the Symptom, Not the Guess 06Make It Argue Against Its Own Plan 07Hand Off to Your Next Session 09Run to the gate, then stop — not step-by-step 10A CLAUDE.md skeleton — turn a description into operating law 11The weekly review prompt — five minutes that turn experience into skill 12Split the work into focused slices, each with clean context 13Is this even worth an agent — ask before you open the chat 14Probe with a small task before you trust it in a new domain

Back to skills

The author

Each story here wraps a lesson paid for in full.

craftagentsomeone building and learning at once

What are you building with agents? Want to trade notes, push back, or build something together — drop a line.

Write hello@craftagent.cloud

58pieces12clustersVI·ENbilingual

✕ Rubber stamp

✓ Real review

Get new pieces by email