agentleFS
Sign inStart free
aniketshukla1aniketshukla1/sdlc-kit/.cursor/skills/mr-feedback/SKILL.md

mr-feedback

Skill1 starsChanged 13 days agoMITView on GitHub

Phase 8. Fixes failed pipeline jobs and addresses MR review comments (GitLab or GitHub), then pushes and posts a summary note. Use when an MR exists and the human says "Continue <KEY>" after review or a pipeline run.

  • Commits and pushes

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, 1 is distinctive, and 5 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.

  • 5

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read PROGRESS.md (branch, MR IID).

    line 12
  2. Very specific:

    After fixes: /run-checks → ALL PASS.

    line 31
  3. Very specific:

    Commit: git commit -m "<KEY>: fix(<scope>): address review feedback" (explicit git add <files>).

    line 32
  4. Very specific:

    Write Agent update: <one line per comment: what you changed or your answer> to .agent/work/<KEY>/note.md, then mr.sh note <IID> …

    line 37
  5. Distinctive:

    Do not resolve threads. The reviewer resolves them.

    line 39
  6. Very specific:

    Update PROGRESS.md: Phase: 8-feedback (round <n> done) + log.

    line 40

≈ “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.

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

MR feedback (phase 8)

You never merge. After a human merges, the pipeline deploys to the Testing environment. mr.sh below is bash .cursor/skills/mr-feedback/scripts/mr.sh: it talks to GitLab or GitHub, whichever hosts the repo. Staging and Production are never your concern.

Steps

  1. Read PROGRESS.md (branch, MR IID). Then: git switch <branch> && git pull --ff-only.
  2. Pipeline and conflicts: bash .cursor/skills/mr-feedback/scripts/pipeline_failures.sh <IID> It exits 1 when it can't read GitLab or GitHub: show that error to the human; never treat it as green. MERGE CONFLICTS → first git fetch origin && git merge origin/main. Conflict in a file you did not change, or you're unsure → git merge --abort and ask the human. Else resolve, git add <files>, git commit --no-edit. Never rebase here: the branch is pushed, so a rebase needs a force push. 2 Alembic heads after the merge → /db-migration step 7. For each failed job, classify it and act:
    ClassSignsAction
    test / lint / typeassertion, ruff, mypy, eslint, tsc errorsReproduce with /run-checks, fix the code
    buildinstall/compile/docker build errorsFix only if caused by your change, else ask
    flaky / infratimeout, runner error, network, image pullmr.sh retry <job-id> ONCE. Fails again → tell the human
  3. Review comments: bash .cursor/skills/mr-feedback/scripts/mr_comments.sh <IID> It shows unresolved threads and comments posted since your last Agent update note (step 6). Make a table in PROGRESS.md: comment | action (fix / answer / needs new ticket).
    • Fix requested changes that are inside the plan scope.
    • Out-of-scope requests: don't do them; answer that it needs a new ticket.
    • Disagree? Don't argue in code. Write your reason in the reply and let the reviewer decide.
  4. After fixes: /run-checks → ALL PASS.
  5. Commit: git commit -m "<KEY>: fix(<scope>): address review feedback" (explicit git add <files>). Push: git push (never force). Refresh the proof for the new head: run proof_pack.sh again (/gitlab-mr step 7), replace the Proof section of .agent/work/<KEY>/MR.md with it, then mr.sh set-body <IID> .agent/work/<KEY>/MR.md.
  6. Post ONE summary note on the MR. Start it with Agent update: next round, step 3 skips what it answers. Write Agent update: <one line per comment: what you changed or your answer> to .agent/work/<KEY>/note.md, then mr.sh note <IID> .agent/work/<KEY>/note.md. Do not resolve threads. The reviewer resolves them.
  7. Update PROGRESS.md: Phase: 8-feedback (round <n> done) + log.
  8. Tell the human what changed and that the MR is ready for another review.

Done when

  • Merge conflicts merged (never rebased) or reported; every failed job fixed, retried once (flaky), or reported
  • Every review comment has an action in PROGRESS.md and a line in the MR note
  • ALL PASS shown, changes pushed without force
  • The MR description's Proof section is refreshed for the pushed head

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.

Check your setup, free