Review
SkillDev toolsPerform a code-review pass focused on bugs and ship readiness. Use for "review this", "before I ship", "code review", and PR checks.
Available today. Use it from your connected AI after setup.
No other account needed.
Connect ahel once, and every AI you use reads what you have installed.
Then ask your AI: use the Review skill
What this skill tells your AI
The instructions your AI receives, as published by uss-parks/lamprey-harness in skills/review/SKILL.md and read by ahel’s review.
Use this skill when the user asks for review rather than implementation.
Inspect the diff and nearby code before judging. Focus on correctness, security, data loss, concurrency, permissions, performance cliffs, and missing tests. Prefer real findings over style commentary. Cite file and line references for every actionable issue.
Use workspace_context when repository state or likely checks are unknown. Use read-only searches freely. Do not edit files unless the user changes the request from review to fix.
Before the verdict, list the failure modes or risks you checked and the evidence consulted for each one: files, diffs, receipts, contracts, tool metadata, or commands actually observed. Also name any unchecked gaps. If no issues are found, say that clearly without inventing one, but still show what you checked.
Output findings first, ordered by severity. For each finding, include severity, file/line, what can go wrong, and the smallest fix.
End with one verdict line: SHIP when no blocking issues remain, or CHANGES when fixes are required.
Stop when the user has enough evidence to merge, revise, or ask for fixes.
Signals
- GitHub stars
- 62
- Forks
- 2
- Last commit
- Sep 2026
- Hacker News mentions
- 20
Advanced
- Catalog kind
- skill
- Gateway key
review-uss-parks- Source
- github.com/uss-parks/lamprey-harness