requesting-review
Skill1 starsChanged 21 days agoMITView on GitHub
Review a unit commit or a whole change against two checks - does it solve the stated problem, and does it follow the project's code style and instructions. One reviewer per unit; two in parallel for the whole change. Reviewers read code and evidence; they do not run specs or lint. Bounded by a findings ledger and a two-round budget.
2 other projects use a version of this.
Adopted by
Most recently taken up by
- MorganOnGitHub/tackletheir own version, first seen Sep 27, 2026
- renathoaz/arsyntheir own version, first seen Mar 24, 2026
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 2
- renathoaz/arsyn★ 2same name
- MorganOnGitHub/tackle★ 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 repository10 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- brainstormingSkill
- development-workflowSkill
- exploringSkill
- finishing-developmentSkill
- systematic-debuggingSkill
- test-driven-developmentSkill
- writing-plansSkill
- writing-skillsSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 34
Write each brief from references/reviewer-brief.md: the plan section, the target and base SHAs, the implementer's report, the checks to …
- Distinctive: line 39
Do not give a reviewer your conclusions, the other reviewer's findings, or hints about what to ignore.
- Very specific: line 46
Round 2+: - only the reviewer who raised a blocking finding comes back.
≈ “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
Requesting Review
Review is a blocking gate on one immutable commit, run by fresh
reviewers. Every review applies two checks:
- Problem check (
references/problem-check.md)- does the work solve the problem described for the project or the unit, as intended?
- Quality check (
references/quality-check.md)- does the work follow the project's code style and instructions, and is it clean and safe?
Reviewers do not run specs or lint. The implementer ran them once at the end of its round and recorded the output in its report. Reviewers read that evidence and judge whether it is enough.
Who reviews what
| Target | Reviewers |
| A unit commit | One reviewer, applying both checks. |
The whole change (finishing-development) | Two reviewers in parallel against the same commit - one per check, so each goes deep. |
Reviewers only read, so parallel reviewers need no coordination.
Dispatch
Write each brief from references/reviewer-brief.md:
the plan section, the target and base SHAs, the implementer's report, the checks
to apply, the round, and - from round 2 - the findings ledger
(references/findings-ledger.md).
Do not give a reviewer your conclusions, the other reviewer's findings, or hints about what to ignore. The ledger is allowed: it records what is settled, not what to think.
Rounds
- Round 1 - the full diff against both checks. Raise everything here.
- Round 2+ - only the reviewer who raised a blocking finding comes back. It
confirms its findings are fixed and reviews the fix diff. A new finding outside
the fix diff is allowed only at
CriticalorImportant. - Budget: two fix rounds. Then stop, report what is open and why the loop is not converging, and ask the user. Do not start a third round on your own.
Verdicts and severity
Each review returns APPROVED, APPROVED_WITH_MINOR, CHANGES_REQUIRED, or
BLOCKED. Each finding has a severity, file:line, and the concrete risk.
CriticalandImportantblock. Collect all blocking findings into one fix brief.Minornever blocks. It goes to the ledger asdeferredand is settled infinishing-development.BLOCKEDmeans the reviewer lacks evidence - for example, the report does not show a required check ran. Get the evidence from the implementer, then re-dispatch. A blocked review is never an approval.
Severity follows risk to correctness, security, or the user - never how easy the
fix is. Comment, wording, and documentation findings are Minor.
Save each review at its assigned path and record the verdict in the ledger and
PROGRESS.md.
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.