agentleFS
Miyamura80Miyamura80/MCP-Template/.agents/skills/push-release/SKILL.md

push-release

Skill1 starsChanged 5 months agoMITView on GitHub

Cut a new release - bump the version in pyproject.toml, tag, and push to trigger PyPI publish and MCP Registry publish via GitHub Actions. Use when the user asks to release, ship a version, bump and tag, cut a hotfix, or publish to PyPI.

  • Installs packages
  • Commits and pushes

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

8 rules. About 0 are common, 7 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    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:

    Make sure main is clean and CI passes locally:

    line 16
  2. Distinctive:

    If anything fails, stop and report.

    line 24
  3. Distinctive:

    Ask the user which bump if not specified:

    line 28
  4. Distinctive:

    Only relevant if PyPI publishing has never been configured for this repo.

    line 93
  5. Distinctive:

    Create the package on pypi.org.

    line 95
  6. Distinctive:

    Add a trusted publisher under the package settings:

    line 96
  7. Very specific:

    Create a release environment in GitHub (Settings -> Environments).

    line 101
  8. Distinctive:

    After this is done once, every tag push auto-publishes.

    line 103

≈ “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

Push Release

This project ships one wheel to PyPI containing all three console scripts (mymcp, mymcp-api, mymcp-mcp) at a single version. One tag push, one PyPI publish - there is no separate PyPI step for the API or MCP. Downstream users pip install the package and run whichever entry point they want; the API and MCP server aren't deployed by this repo. The MCP Registry entry refreshes automatically on the same tag (mcp-registry-publish.yml), so it's discovery metadata, not a second distribution.

Workflow

Execute these steps in order. Confirm with the user before running make bump_version and before pushing - tags are public and a published version cannot be unpublished from PyPI.

1. Pre-flight

Make sure main is clean and CI passes locally:

git status                # working tree must be clean
git rev-parse --abbrev-ref HEAD   # must be on main
make ci && make test

If anything fails, stop and report. Do not proceed with a broken main.

2. Pick the bump level

Ask the user which bump if not specified:

  • BUMP=patch (default) - bug fixes, no API change. 0.1.1 -> 0.1.2
  • BUMP=minor - additive features, backwards compatible. 0.1.1 -> 0.2.0
  • BUMP=major - breaking changes. 0.1.1 -> 1.0.0

3. Bump the version

make bump_version BUMP=patch   # or minor / major

This:

  • updates pyproject.toml version = "x.y.z"
  • creates a Release vX.Y.Z commit
  • creates an annotated tag vX.Y.Z

The Makefile target refuses to run if the working tree or staging area is dirty.

4. Push with tags

git push origin main --follow-tags

The tag push is what triggers the release - without --follow-tags only the commit is pushed and nothing publishes.

5. Verify

Two GitHub Actions workflows fire on the v* tag:

  1. Release (.github/workflows/release.yml)
    • Runs make ci and make test
    • Builds the package with uv build
    • Publishes to PyPI via OIDC trusted publishing
    • Creates a GitHub Release with auto-generated notes, wheel, and sdist
  2. MCP Registry Publish (.github/workflows/mcp-registry-publish.yml)
    • Authenticates via GitHub OIDC
    • Publishes the MCP server to the MCP Registry

Check:

  • GitHub Actions tab - both workflows green
  • GitHub Releases - new vX.Y.Z entry with assets
  • MCP Registry - updated server entry

Hotfix release

For an urgent fix on top of an existing release without picking up unrelated main changes:

git checkout -b hotfix/<description> v0.1.1
# apply the fix
make ci && make test
# open a PR to main, squash-merge it
git checkout main && git pull
make bump_version BUMP=patch
git push origin main --follow-tags

Version is single-sourced

The version lives in one place: pyproject.toml -> version = "x.y.z". All three interfaces read it at runtime via importlib.metadata.version(). Never hand-edit version strings in source files.

PyPI trusted publishing (one-time setup)

Only relevant if PyPI publishing has never been configured for this repo. The release workflow uses PyPI trusted publishing, so no API tokens are needed:

  1. Create the package on pypi.org.
  2. Add a trusted publisher under the package settings:
    • Owner: GitHub username or org
    • Repository: GitHub repository name
    • Workflow: release.yml
    • Environment: release
  3. Create a release environment in GitHub (Settings -> Environments).

After this is done once, every tag push auto-publishes.

What can go wrong

  • Working tree not clean - make bump_version aborts. Commit or stash first.
  • Pushed commit without tag - nothing publishes. Push the tag separately: git push origin vX.Y.Z.
  • CI fails inside release.yml - the PyPI publish step never runs, but the tag exists. Fix the cause on main, then either bump again to a new patch or delete the tag (git push --delete origin vX.Y.Z) and re-run after fixing - delete only if no one has pulled it.
  • PyPI publish succeeds but MCP Registry fails - the version is live on PyPI; fix the registry workflow and re-run it via workflow dispatch.

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.