code-review-pass
Skill9 starsChanged 2 months agoMITView on GitHub
Reviews a diff, branch, or pull request across correctness, security, architecture, readability, and performance, returning ranked findings with file and line references and a concrete fix for each. Use before merging, when checking code written by someone or something else, or when a change needs a quality gate.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- youssef-tharwat/code-review-passtheir own version, first seen Sep 28, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- youssef-tharwat/code-review-pass★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository28 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- api-designSkill
- architecture-decisionSkill
- autonomous-relaySkill
- browser-verificationSkill
- data-analysisSkill
- data-modelingSkill
- data-pipelineSkill
- data-qualitySkill
- documentationSkill
- firmware-buildSkill
- frontend-buildSkill
- incident-responseSkill
- incremental-deliverySkill
- ml-lifecycleSkill
- performance-passSkill
- release-and-rollbackSkill
- requirements-interviewSkill
- security-hardeningSkill
- simplification-passSkill
- spec-writingSkill
- systematic-debuggingSkill
- tdd-loopSkill
- team-escalationSkill
- using-agent-dev-teamSkill
- work-breakdownSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 50
Name one specific thing done well: Not flattery: it shows you read the change.
- Distinctive: line 62
Do not inflate severity to be heard.
- Very specific: line 95
Every finding has a file:line and a concrete fix.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Code Review Pass
A review is finished when a person can act on every finding without asking you a follow-up question.
Process
- Get the actual diff. Reviewing from a description reviews the description. If you cannot see the change, ask for it.
- Read the tests first. They state the intended behavior. If they do not tell you what the change is supposed to do, that is finding number one.
- Read enough context around each change to know what the surrounding code assumes. A changed line is not reviewable alone.
Review along five axes, in this order. Order matters because it determines what you still have attention for.
Correctness — Does it do what it claims? Empty, one, many, boundary, error path, concurrent, and the retry. Do the tests assert the behavior, or do they assert that the implementation is the implementation?
Security — Untrusted input, server-side authorization, injection sinks, secrets, error messages that leak. See
references/security-checklist.md. Anything on a trust boundary routes tosecurity-hardeningfor depth.Architecture — Does it fit the patterns already here? Is the dependency direction right? Is a boundary being punched through for convenience?
Readability — Naming, control flow, function size. A comment explaining what the code does is usually a name that should have been better.
Performance — N+1, work in a loop that belongs outside it, an unbounded query, a missing timeout.
- Rank by leverage. Lead with what would hurt in production. A style note above a missing authorization check makes the whole review skimmable.
- Name one specific thing done well. Not flattery: it shows you read the change.
- Give a fix with every finding. "This is fragile" is not actionable.
Severity
| Severity | Meaning | Merge |
| Critical | Data loss, auth bypass, remote code execution, corruption | Blocks |
| Important | Missing test for new behavior, wrong error handling, architectural violation | Fix before merge |
| Suggestion | Readability, naming, optional simplification | Author's call |
Do not inflate severity to be heard. Do not pad with suggestions to look thorough.
What to look for that tools do not catch
- A test that asserts the implementation rather than the behavior.
- Error handling that catches and continues, leaving the system in a half-state.
- A new code path with no test, hidden inside a diff that has many tests.
- An interface change with a caller in the repository that was not updated.
- A comment that no longer matches the code beneath it.
- A migration with no rollback.
- Authorization checked in the handler for one path and not for the sibling path added in the same diff.
Output
## Review: <target>
**Verdict:** approve | approve with changes | request changes
**Strength:** <one specific thing done well>
### Critical
- `path:line` — <problem>. <fix>.
### Important
### Suggestions
### Not reviewed
<what you could not see, and why it matters>Verification
- Every finding has a
file:lineand a concrete fix. - Findings cover more than one axis, or you say why the others did not apply.
- The verdict matches the findings: an open Critical means request changes.
- You state what you did not review. A silent gap reads as coverage.
Red flags
| Thought | Reality |
| "Looks good to me" | Say what you checked and what you could not. LGTM is not a review. |
| "I'll fix it while I'm here" | Then nobody reviewed it. Report, do not edit. |
| "Twenty findings" | Rank them, or the Critical gets skimmed past. |
| "Tests pass, so correctness is covered" | Passing tests prove the tests pass. Read the assertions. |
| "Big diff, I'll spot-check" | Say you spot-checked and which parts. An implied full review is worse than an honest partial one. |
| "The author is senior, it is probably fine" | Then the review is quick. It is not optional. |
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

