agentleFS
seancabalseseancabalse/agentic-playbook/skills/ship-pr/SKILL.md

ship-pr

Skill0 starsChanged 15 days agoMITView on GitHub

Run the checks, update the docs, commit and open the PR from the template. Run only when the user asks to ship.

31 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use ideate, review-pr and 6 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.
  • OpenAISkill

    ideate

    Generate image-based alternatives, remixes, or new design directions from a Product Design brief. Use when the user asks for design variants, visual exploration, remixes, or image-generated approaches from provided context.

    Also in 4 projects that use this

  • DockerSkill

    review-pr

    Review one or more incoming Docker documentation pull requests as a maintainer. Independently validate technical claims, assess editorial fit and information architecture, choose a verdict, and draft exact inline or PR-wide feedback behind a confirmation gate. Use for requests such as "review PR 123", "is this PR correct?", "does this information belong here?", "validate this PR", or "help review backlog PRs". Do not use to maintain or fix a PR you own; use maintain-pr for that.

    Also in 4 projects that use this

  • Google CloudSkill

    triage

    Gather context on a change or bug and plan the implementation.

    Also in 4 projects that use this

  • GoogleSkill

    diagnosing-bugs

    Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.

    Also in 3 projects that use this

  • GoogleSkill

    domain-modeling

    Build and sharpen a project's domain model. Use when the user wants to pin down domain terminology or a ubiquitous language, record an architectural decision, or when another skill needs to maintain the domain model.

    Also in 3 projects that use this

  • +3 more from google

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 31 edited versions we read keep these.

  • 6

    Distinctive

    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
  1. Distinctive:

    Run the "before calling a task done" commands from AGENTS.md: lint, typecheck, unit tests, plus DB and integration tests if data, policies, …

    line 16
  2. Very specific:

    Read the full diff as a reviewer would (or run the code-reviewer agent):

    line 20
  3. Distinctive:

    Every changed line traces to the task.

    line 22
  4. Distinctive:

    No debug code, test-only hacks, hard-coded colours or strings, or secrets.

    line 23
  5. Very specific:

    Run docs-upkeep: AGENTS.md rules, Decision Log, Architecture, runbook traps, README if setup changed.

    line 28
  6. Distinctive:

    If they're missing, stop and capture them.

    line 41
  7. Very specific:

    Only if nothing a person sees changes (a refactor, a type-only or test-only change), write Screenshots: none — <why nothing visible …

    line 42
  8. Distinctive:

    No AI attribution in the PR body either.

    line 55
  9. Distinctive:

    Stop processes you started for this task.

    line 61

“Common” is measured: kept by at least 6 in 10 of the 31 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

Ship it

Playbook files named here: when the project has its own copy at .playbook/local/<same path>, use that copy instead.

Ticket: $ARGUMENTS

1. Checks (all must pass; fix before continuing)

Run the "before calling a task done" commands from AGENTS.md: lint, typecheck, unit tests, plus DB and integration tests if data, policies, functions or sync changed. Run the generated-types check if the schema changed. Paste failures; don't summarise them away.

2. Self-review

Read the full diff as a reviewer would (or run the code-reviewer agent):

  • Every changed line traces to the task. Nothing speculative. Nothing you orphaned is left behind (imports, files, copy keys).
  • No debug code, test-only hacks, hard-coded colours or strings, or secrets.
  • Unhappy paths are handled; the copy is in the catalogue; accessibility is covered.

3. Docs (same PR)

Run docs-upkeep: AGENTS.md rules, Decision Log, Architecture, runbook traps, README if setup changed.

4. Commit

  • Stage only the files you meant to change (git status first; never git add -A blindly).
  • The message is an imperative subject with the ticket id, then a body explaining what changed for the user and why.
  • No Co-Authored-By, no "Generated with", no AI mention, even if a tool default suggests one.

5. Screenshots gate

List what the branch changes: git diff --name-only <base>...HEAD. If any change can alter what a person sees (components, screens, pages or routes, styles and design tokens, copy in the message catalogue, images and icons, email, notification or PDF templates), the PR needs screenshots:

  • They must exist before you open the PR: screenshots/<ticket>/, from capture-screenshots (before and after, plus the states and settings that matter, per principles/04-design-and-ux.md #17).
  • If they're missing, stop and capture them. If the app can't run here (a device, an account or data you don't have), stop and ask the owner; don't open the PR without them unless the owner says so in this conversation.
  • Only if nothing a person sees changes (a refactor, a type-only or test-only change), write Screenshots: none — <why nothing visible changes> in the PR body instead.

6. PR

If AGENTS.md section 8 says Delivery: local merge, there is no PR. Do the first bullet below, but do not push. Write the title and the description as below. Then give the owner the commands to squash-merge the branch into its base: git checkout <base>, git merge --squash <branch>, and git commit with the title as the subject and the description as the body. Do not merge yourself.

  • Bring the branch up to date with sync-branch (rebase onto the team branch, checks, --force-with-lease). Then push the branch, and open the PR against the default branch (or the stacked base, if its PR is still open; check it isn't already merged).
  • The title becomes the commit on the team branch (PRs are squash-merged, principles/10-delivery-ci-release.md #3): imperative, customer-facing, with the ticket id. Don't squash the branch yourself.
  • Fill the body from ${CLAUDE_PLUGIN_ROOT}/templates/pr-description.md:
    • Verified: actual results.
    • Not verified: honestly stated.
    • Needs: new build, env var, owner action.
    • Screenshots for UI changes, from the gate in section 5. Publish them with ${CLAUDE_PLUGIN_ROOT}/templates/testing/push-screenshots.sh <PR> <files…> (or the project's own script) to a screenshots branch, never the default branch. Embed the lines it prints, and check each image loads.
  • No AI attribution in the PR body either.

7. Afterwards

  • Move the ticket to the tracker's review status, link the PR the way that tracker links them, and comment with what was done and how it was verified (a handoff note). See AGENTS.md → "Tracker".
  • Watch CI. If it fails, fix it before starting other work.
  • Stop processes you started for this task.
  • Tell the owner: the PR link, what's verified, what isn't, and anything they need to do. Merging is theirs, with Squash and merge.

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.