agentleFS
rpraharajrpraharaj/ai-engineer/skills/review-change/SKILL.md

review-change

Skill1 starsChanged 17 days agoMITView on GitHub

Review a local or branch diff for bugs and engineering-bar failures. Use when the user wants a code review, review of uncommitted work, or a second look at a branch diff. Write structured findings. Do not edit product code. Do not post to a git host unless asked and a host CLI works.

29 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

Used with

Projects with it also use plan-change, write-runbook and 2 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • BanKnightSkill

    plan-change

    将已完成 spec/design 的 change 拆成可执行 plan 与 tasks。用户要把设计落成执行计划、任务清单或实现步骤时使用。

    Also in 7 projects that use this

  • MaheshAwasareSkill

    write-runbook

    Generate an operational runbook for a service from its code, config, and dependencies — what to do when it breaks, where to look, how to roll back, who to escalate to. Use proactively before a service goes to production, after a major incident, or when handing off operational ownership.

    Also in 3 projects that use this

  • ShopifySkill

    prepare-release

    Prepare the changeset-release/main branch for release by syncing version.ts files and running pnpm install

    Also in 3 projects that use this

  • Microsoft AzureSkill

    release-notes

    Author and update release notes for Azure App Configuration libraries and providers. USE FOR: write release notes, add release note entry, update release notes, draft release notes, new version release notes, format release notes. DO NOT USE FOR: generating changelogs from git history, writing internal engineering notes.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

12 rules. 0 are common, 8 are distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 25 edited versions we read keep these.

  • 8

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read-only review of a diff.

    line 12
  2. Distinctive:

    If the diff is empty, say so and stop.

    line 20
  3. Distinctive:

    Read the changed files, not only the hunks, when a finding depends on callers or types.

    line 22
  4. Very specific:

    If a shape-work document, spec, or ticket the user pointed at exists, report it separately from Issues (do not fold it into bug/nit):

    line 37
  5. Very specific:

    If there is no spec, write Spec: n/a and skip.

    line 43
  6. Distinctive:

    Do not flood nits when there are bugs.

    line 45
  7. Distinctive:

    Write findings in the reply.

    line 53
  8. Distinctive:

    If there are no issues, keep Summary and an empty Issues section.

    line 74
  9. Very specific:

    Only if the user asked to post the review and gh or glab is authenticated:

    line 80
  10. Distinctive:

    Put the rest in the review body

    line 83
  11. Very specific:

    Do not submit/approve/request-changes unless the user named that action

    line 84
  12. Distinctive:

    If the CLI is missing, print the findings and stop.

    line 86

“Common” is measured: kept by at least 6 in 10 of the 25 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Review change

Read-only review of a diff. Do not edit product source. Do not commit.

Named review-change so it does not replace a host's own review command.

Target

Load engineering-bar and follow its Shared procedures for change scope and evidence. Honor an explicit PR/MR or branch target; otherwise include branch commits and relevant local edits, including untracked files. If remote access is unavailable, request a local revision or diff. Preserve the same scope in specialist checks.

If the diff is empty, say so and stop. If it is huge (multi-megabyte or hundreds of files), summarize by file and ask what to narrow; do not pretend you reviewed it all.

Read the changed files, not only the hunks, when a finding depends on callers or types.

Load the bar

Load engineering-bar (paths in that skill). Use it for hygiene, structure, failure, data, security, performance guards, and operability.

Also look for:

  • Incorrect behavior vs the stated intent
  • Missing tests for new logic (point at write-tests; do not write them here)
  • Authz / injection / secrets if the diff touches those (point at security-pass for a full pass)
  • Over-engineering only (point at lean-review; do not turn this review into a delete-list)

Spec axis

If a shape-work document, spec, or ticket the user pointed at exists, report it separately from Issues (do not fold it into bug/nit):

  • Missing or partial vs the acceptance checks
  • Behavior in the diff that was not asked for
  • Checks that look implemented but look wrong

If there is no spec, write Spec: n/a and skip. Do not invent a spec.

Do not flood nits when there are bugs. Severity:

  • bug — wrong, unsafe, or will fail in production
  • suggestion — real improvement, not blocking unless the user asked for a hard bar
  • nit — local style; skip if the repo's formatter would handle it

Findings file

Write findings in the reply. Also write a file only if the user asked, or if this repo already stores reviews.

## Summary
<2–4 sentences: what the change does, whether it looks correct, main risks>

## Issues

### Issue 1 — Severity: bug
- File: path/to/file.ext:LINE
- Description: <what is wrong>
- Suggestion: <how to fix>
- Status: open

## Bar
<engineering-bar fails and waivers only>

## Spec
<missing / extra / wrong, or n/a>

If there are no issues, keep Summary and an empty Issues section. Do not invent findings.

Line numbers refer to the new file. If you cannot pin a line, say so and put the finding in Summary.

Optional publish

Only if the user asked to post the review and gh or glab is authenticated:

  • Map each issue with a valid new-file line onto an inline comment
  • Put the rest in the review body
  • Do not submit/approve/request-changes unless the user named that action

If the CLI is missing, print the findings and stop.

Done

Issue counts by severity, the summary, and the path of any file you wrote. Do not implement fixes unless the user asked — that is implement-change or debug.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.