agentleFS
saleorsaleor/saleor-dashboard/.claude/skills/saleor-dashboard-pr-quality-check/SKILL.md

saleor-dashboard-pr-quality-check

Skill1.1k starsChanged 8 days agoBSD-3-ClauseView on GitHub

Check a PR (or local branch when no PR is open) against the repo's PR rules — one significant change per PR, a title and description that explain the whole scope, before/after screenshots for UI. Proposes a title and description, offers screenshots and a `gh pr edit` update, and plans a split into smaller PRs with hand-off prompts for other agents. Use when planning work, after each commit, before opening or updating a PR, or when asked to "check the PR", "PR quality", "review PR description", or "should I split this PR".

  • Commits and pushes

Trending

Less talk lately.about the whole repository

Dec 2025: 0 mentionsJan 2026: 0 mentionsFeb 2026: 3 mentionsMar 2026: 5 mentionsApr 2026: 0 mentionsMay 2026: 1 mentionJun 2026: 33 mentionsJul 2026: 3 mentionsAug 2026: 0 mentionsSep 2026: 0 mentionsOct 2026: 2 mentionsNov 2026: 0 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the saleor/saleor-dashboard repository that our weekly crawl found. There is no adoption data for this file yet.

Who's talking about saleor/saleor-dashboard

The biggest 5 of 50 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 50

Talked about on GitHub (50 times).

On GitHub · 50

45 more on GitHub

The file

We split this skill into the rules it gives an agent

10 rules. About 0 are common, 4 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

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

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

    Read .github/CONTRIBUTING.md under Pull requests for scope policy and .github/PULL_REQUEST_TEMPLATE.md for title and description …

    line 8
  2. Very specific:

    If there is no PR (or gh is unavailable), use what exists locally:

    line 24
  3. Very specific:

    Read .changeset/*.md files added on the branch.

    line 35
  4. Distinctive:

    When there is more than one significant change, decide how they relate:

    line 49
  5. Distinctive:

    Put the scope finding first in your report, not in a footnote.

    line 58
  6. Very specific:

    Write a ready-to-paste title and description that pass step 3.

    line 77
  7. Very specific:

    Remove optional sections only as the template permits.

    line 81
  8. Distinctive:

    When step 2 found more than one significant change, propose a split:

    line 101
  9. Very specific:

    Write the plan to .context/pr-split-plan.md when that directory exists.

    line 113
  10. Distinctive:

    Keep it short, most important first:

    line 119

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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

PR quality check

Read .github/CONTRIBUTING.md under Pull requests for scope policy and .github/PULL_REQUEST_TEMPLATE.md for title and description requirements. These files are the sources of truth for the checks below.

This skill only reads and proposes. Every outward action (editing the PR, running the app for screenshots, creating branches) needs the user's explicit confirmation. Never push.

1. Gather the change

Find an open PR for the current branch:

gh pr view --json number,url,title,body,baseRefName,files,commits 2>/dev/null

If there is no PR (or gh is unavailable), use what exists locally:

git log --format='%h %s' origin/main..HEAD
git diff --stat origin/main...HEAD
git status --short

Include uncommitted changes. During planning, when nothing is written yet, check the plan itself: its goal and the list of changes it intends to make.

Read .changeset/*.md files added on the branch; they state the intended user-visible change.

2. Check scope

Classify every change into one kind: bug fix, refactor, feature / behavior change, dependencies, tooling / CI, tests only, docs, generated (GraphQL types, lockfile, messages — these follow their source and do not count on their own).

Count significant changes. A change is significant when a reviewer would have to evaluate it on its own merits. Supporting edits that exist only to make the main change work do not count.

When there is more than one significant change, decide how they relate:

  • Unrelated changes, or a change hidden inside another (for example a functional change in a refactor commit) → blocker. They must be split. Go to step 6.
  • Connected changes (for example a refactor that enables the fix) → strongly recommend splitting and go to step 6. Say it is acceptable to keep them together only if each commit holds just its own change.

Put the scope finding first in your report, not in a footnote.

3. Check title and description

Against the PR template requirements, report each as pass / missing / weak:

  • Title says what changes; user-facing wording for product changes, conventional prefix for internal ones.
  • Problem, change (covering every area in the diff), non-obvious decisions, screenshots for UI, testing, links.
  • The description mentions everything the diff touches. Flag any changed area the description does not mention — that is the most common gap.
  • The PR template checkboxes are addressed.

Locally (no PR), check the commit messages and changesets the same way; they will become the PR text.

4. Propose title and description

Write a ready-to-paste title and description that pass step 3. Base it on the diff, not on the existing text. Follow the headings and instructions in .github/PULL_REQUEST_TEMPLATE.md; do not prescribe a separate format.

Remove optional sections only as the template permits. Keep its checkboxes at the end. Leave a clear TODO for anything you could not determine instead of inventing it.

5. Offer screenshots and a PR update

Screenshots — when the diff touches UI (src/**/*.tsx, *.module.css, messages), propose taking before and after screenshots of the affected screens. Name the screens and the state to capture. Ask for confirmation before starting the app; it needs a backend and memory. Take "before" on origin/main and "after" on the branch, save them under .context/, and show them to the user.

Update the PR — when a PR exists, offer to apply the proposed title and description with gh pr edit <number> --title ... --body-file <file>. Show the exact text first and ask for confirmation. Do not update without it. When no PR is open, offer the text for gh pr create instead, and do not create it unless asked.

6. Split plan and agent hand-off

When step 2 found more than one significant change, propose a split:

  1. One entry per future PR: title, kind, the files or hunks it takes, and which commits map to it.
  2. Order: prerequisites (refactors, dependency bumps) first; state which PR stacks on which, and whether each can merge independently.
  3. For each PR, a self-contained hand-off prompt another agent can run in its own workspace: base branch, what to take from this branch (git checkout <branch> -- <paths> or specific commits), what to leave out, required checks from AGENTS.md → Verification, changeset yes/no, and the proposed PR title and description.

Write the plan to .context/pr-split-plan.md when that directory exists; otherwise return it in the conversation. Ask for confirmation before creating any branch or starting agents.

Report format

Keep it short, most important first:

  1. Scope: single change ✅, ⚠️ split recommended (connected), or ⛔ split required (unrelated or hidden) — list the changes found.
  2. Title / description: failing items only.
  3. Proposed title and description.
  4. Questions: screenshots? update the PR? create the split plan?

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.