agentleFS
ChariereFiedlerChariereFiedler/ronce-racine/skills/merge-request-review/SKILL.md

merge-request-review

Skill9 starsChanged 25 days agoMITView on GitHub

Use when reviewing a merge/pull request or a branch diff before merge - "review this MR", "review this PR", "validate this branch diff", "review cette MR", "review cette PR", "valide ce diff de branche", before integrating someone else's work into the mainline.

  • Reads credentials

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 3 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  • could be automatic: a setting could enforce it
  1. Distinctive:

    This skill: when: reviewing the diff of an MR/PR or a branch that is someone else's or not yet integrated before merge - "review this PR", …

    line 16
  2. Very specific:

    commit-readiness-review: instead if: it is my own changes to commit (before commit, not before merging a third-party branch)

    line 17
  3. Very specific:

    superpowers:requesting-code-review: instead if: I want my own finished work reviewed, not to review someone else's

    line 18
  4. Very specific:

    validating-features-end-to-end: instead if: the question is "does it work" (functional proof), not "is it mergeable"

    line 19
  5. Distinctive:

    Read the whole diff first, then work through the checklist:

    line 35
  6. Distinctive:

    If there is a conflict with the target: rebase/merge the target into the source, re-validate lint/test, re-push - never force-push the …

    line 50

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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.

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

Merge Request Review - reviewing a change before integration

If the current repo has a specific review skill, it wins - it knows the forge, the tracker, the API commands and the project's optional auto-merge.

This skill vs. others

  • This skill when: reviewing the diff of an MR/PR or a branch that is someone else's or not yet integrated before merge - "review this PR", "validate this diff", "OK to merge?"
  • commit-readiness-review instead if: it is my own changes to commit (before commit, not before merging a third-party branch)
  • superpowers:requesting-code-review instead if: I want my own finished work reviewed, not to review someone else's
  • validating-features-end-to-end instead if: the question is "does it work" (functional proof), not "is it mergeable"

Principle

  • Approving = having read the diff line by line. A "green pipeline" proves the existing tests pass, not that the code is correct, safe or complete.
  • The review produces an explicit verdict: approve / request changes - never a vague "looks good".

Context to gather (before acting)

  • The full diff of the source branch against the target: git fetch <remote> <source-branch> then git diff <remote>/<target>...<remote>/<source-branch> (not just the forge summary)
  • The pipeline / CI status of the MR/PR (green / red / running) via the forge UI or CLI
  • The linked ticket / issue: what the change is supposed to do → check the diff does it, and only that
  • The repo conventions: message format, lint/typecheck/test per stack (package.json / Cargo.toml / Makefile / CI config)

Protocol

Read the whole diff first, then work through the checklist:

- [ ] Pipeline / CI green (and looked at: which jobs, not just the overall badge)
- [ ] Debug leftovers in added lines (console.log, print, dbg!, debugger, unjustified unwrap/expect)
- [ ] Hard-coded waits (waitForTimeout, sleep) in tests
- [ ] Contract / API respected: public signatures, props, events, schema - no unflagged breaking change
- [ ] i18n: no visible hard-coded string if the project is internationalized
- [ ] Sensitive files: .env, *.pem, *.key, credentials, secrets → STOP
- [ ] Disabled tests (.skip, xit, #[ignore]) without justification
- [ ] No conflict with the target (otherwise → resolve before merge)
- [ ] The diff does what the ticket says, and nothing out of scope
- [ ] Explicit verdict issued: approve OR request changes

If there is a conflict with the target: rebase/merge the target into the source, re-validate lint/test, re-push - never force-push the default branch. Schema migration already applied → do not modify the old one, create a new one.

Templates

  • templates/mr-review-report.md - review report (verdict, blocking points, minor points)

Traps & rationalizations

ExcuseReality
"Pipeline green, I approve"The pipeline tests the existing code, not the correctness or safety of the new code. Read the diff
"Small diff, no need to read"A secret or a breaking change fits in one line
"It's urgent, I merge"Urgency does not change the checklist; a broken merge costs more than a review
"The contract doesn't look touched"Check the public signatures/props/events called elsewhere, don't assume
"Minor conflict, I force-push"Always --force-with-lease, never on the default branch

Exit condition

  • Full diff read (not just the forge summary)
  • Whole checklist worked through; every gap → blocking or justified
  • Verdict issued: approve or request changes, with reasons
  • If merged: green pipeline confirmed and no sensitive files

As long as a blocking point remains, do not approve.

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.