agentleFS
gabrielSantosLimagabrielSantosLima/harness/skills/release-cut/SKILL.md

Release cut

Skill1 starsChanged 8 days agoMITView on GitHub

Trigger the Release workflow, watch it, and verify the tag, release notes and attached artifacts. Use when the user asks to cut, publish or ship a release.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 5 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    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. Very specific:

    Read AGENTS.md and .harness.json.

    line 13
  2. Distinctive:

    If the working tree is dirty, stop and ask whether to commit first.

    line 20
  3. Distinctive:

    Ask the user, in one round:

    line 24
  4. Distinctive:

    Wait a moment, then find the run:

    line 42
  5. Very specific:

    Read the failing job log: gh run view <run-id> --log-failed.

    line 77
  6. Distinctive:

    Never move, delete or reuse a tag that already exists.

    line 79
  7. Distinctive:

    Never force-push to fix a release.

    line 81

≈ “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

Release cut

Cut a release through the project's GitHub Actions workflow. Never create tags or releases by hand.

1. Read the policy

Read AGENTS.md and .harness.json. Confirm:

  • the release trigger is workflow_dispatch in .github/workflows/release.yml;
  • the version source (tags or file);
  • whether build artifacts are attached;
  • the current branch and whether it is clean.

If the working tree is dirty, stop and ask whether to commit first.

2. Ask for the release inputs

Ask the user, in one round:

  1. bump type: patch (default), minor, major, or a pre-release type (prerelease, premajor, preminor, prepatch);
  2. pre-release identifier when a pre-release type is chosen (default alpha);
  3. an explicit version, if the user prefers one (overrides the bump type);
  4. a short release summary, used in the notes title.

3. Dispatch

gh workflow run release.yml \
  -f bump=<bump> \
  -f preid=<preid> \
  -f version="<explicit-version-or-empty>" \
  -f summary="<summary>"

Wait a moment, then find the run:

gh run list --workflow release.yml --limit 1

Watch it to completion:

gh run watch <run-id> --exit-status

4. Verify

git fetch --tags
gh release view <tag>
gh release view <tag> --json tagName,isPrerelease,assets

Confirm:

  • the tag matches vX.Y.Z and was created from the intended commit;
  • the release notes are not empty and match docs/RELEASE_TEMPLATE.md;
  • the version contains - if and only if the release is a pre-release;
  • the expected artifacts are attached when attachArtifacts is true;
  • the version file matches the tag when the version source is file.

5. Report

Report the tag, the release URL, the commit, the prerelease flag, the artifact list and the run result.

On failure

  • Read the failing job log: gh run view <run-id> --log-failed.
  • Report the cause and stop.
  • Never move, delete or reuse a tag that already exists. If the tag is wrong, cut a new patch release with the fix.
  • Never force-push to fix a release.

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.