agentleFS
jacobaraujo7jacobaraujo7/remote_pi/.claude/skills/deploy-cockpit/SKILL.md

deploy-cockpit

Skill475 starsChanged 5 months agoMITView on GitHub

Release the Cockpit desktop/mobile app. Use when the user asks to deploy, release, publish or "criar a tag" of Cockpit, bump its version, or watch/diagnose the cockpit-release GitHub Action. Bumps cockpit/pubspec.yaml and CHANGELOG.md, commits on main, pushes the cockpit-v<version> tag and follows the workflow until every job finishes, fixing and re-tagging on failure.

  • Commits and pushes

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 3 are distinctive, and 1 is 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.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Never kill Cockpit or any user process while releasing.

    line 16
  2. Distinctive:

    Do not add the old "beta of 2.0.0" sentence to changelog sections: 2.0 shipped, the line was removed on 2026-09-27 and the updater shows …

    line 17
  3. Distinctive:

    No em-dashes in commit messages or changelog text.

    line 18
  4. Very specific:

    Keep the first ~20 non-empty lines meaningful on their own (the download page shows only those).

    line 45

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 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

deploy-cockpit

Release of the Cockpit app (cockpit/). Trigger: tag cockpit-v<version> on the repo root, workflow .github/workflows/cockpit-release.yml. It produces macOS dmg, Windows exe, Linux deb/rpm (x64 + arm64), Android apk + aab, SHA256SUMS, latest.json and the Sparkle appcasts, and publishes latest.json to rp-s3 automatically. Tag = publication, there is no manual gate.

Rules that always apply:

  • Work on main. Never create a branch for a release.
  • Never kill Cockpit or any user process while releasing.
  • Do not add the old "beta of 2.0.0" sentence to changelog sections: 2.0 shipped, the line was removed on 2026-09-27 and the updater shows the first section verbatim.
  • No em-dashes in commit messages or changelog text.
  • Report the per-job result to the user; if a job fails, say which and why.

1. Pre-flight

cd cockpit
git status --short            # must be clean, or only the files you are about to commit
git pull --ff-only
flutter analyze && flutter test   # only if code changed since the last green run
grep '^version:' pubspec.yaml     # x.y.z+n
awk '/^## /{print; exit}' CHANGELOG.md

The workflow's meta job fails when the tag version differs from pubspec.yaml or when the first ## section of CHANGELOG.md is not the version being released. Both must be updated before tagging.

2. Bump

Patch bump unless the user says otherwise: 1.28.32+95 -> 1.28.33+96 (version and build number both increase).

  1. pubspec.yaml: version: x.y.z+n.
  2. CHANGELOG.md: add ## [x.y.z] - YYYY-MM-DD at the top, in English, starting with the beta sentence, then ### Added / ### Fixed / ### Changed lists derived from git log <last-tag>..HEAD -- cockpit/. Keep the first ~20 non-empty lines meaningful on their own (the download page shows only those). No ## [Unreleased].
  3. Commit and push:
git add pubspec.yaml CHANGELOG.md
git commit -m "chore(cockpit): bump x.y.z+n"
git push

This skill publishes the app only. It never tags or releases cockpit-server; that is the deploy-server skill, run only when the user asks for it.

3. Tag and watch

git tag cockpit-vx.y.z
git push origin cockpit-vx.y.z
sleep 20
RUN=$(gh run list --workflow cockpit-release.yml --limit 1 --json databaseId --jq '.[0].databaseId')

Watch in the background so the conversation stays free (the run takes 15 to 25 minutes, macOS is the slowest):

gh run watch $RUN --exit-status >/dev/null 2>&1; echo "exit=$?"
gh run view $RUN --json conclusion,jobs --jq '.conclusion, (.jobs[] | "\(.name): \(.conclusion)")'

Jobs: Valida tag ↔ pubspec, cockpit-server (fatia x86_64), macOS dmg, Windows exe, Linux deb+rpm (amd64), Linux deb+rpm (arm64), Android apk + aab, GitHub Release + latest.json. All eight must be success.

4. On failure

gh run view $RUN --json jobs --jq '.jobs[] | select(.conclusion=="failure") | "\(.databaseId) \(.name): " + ([.steps[] | select(.conclusion=="failure") | .name] | join(", "))'
gh run view --job <jobId> --log-failed | tail -80

Fix the cause in code or workflow, commit on main, then release a NEW patch version (bump again, new tag). Do not force-move a cockpit-v* tag that already produced a GitHub Release or a latest.json: the auto-updater may already have seen it. Known past causes: a Linux-only test compiled on Windows, dart build cli --target-os needing a newer Flutter, Zig cross-link rejecting a Rust linker flag, a missing artifact making the 8-asset check fail.

5. After success

Report the version, the run id and the asset list:

gh release view cockpit-vx.y.z --json assets -q '.assets[].name'

Mention what the user should test in this build, and remind that latest.json is served with a 5 minute cache, so the in-app updater offers the release shortly after.

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.