agentleFS
LangChainKnown publisherlangchain-ai/langchain-litellm/.claude/skills/conventional-commits/SKILL.md

conventional-commits

Skill62 starsChanged 3 days agoMITView on GitHub

Use before writing a commit message, PR title, PR description or branch name, or adding or removing a label in langchain-litellm. Covers the Conventional Commits title rules that pr-title-lint.yml checks, the PR description rules that keep release-please from dropping or duplicating a commit, branch names, dependency commits that must ship in a release, and which labels bots own.

62 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use github, grilling and 9 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.
  • NVIDIASkill

    github

    Interact with GitHub using the gh CLI restricted to REST API only (no GraphQL). Use when the user wants to work with GitHub issues, pull requests, repos, releases, or Actions — especially in sandboxed environments where the GraphQL endpoint may be blocked. Trigger keywords - github, gh, pull request, PR, github issue, github actions, workflow, release, gh api.

    Also in 3 projects that use this

  • GoogleSkill

    grilling

    Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.

    Also in 5 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 4 projects that use this

  • NVIDIASkill

    research

    Before designing a feature, research how comparable libraries solve the problem, check best practices, known pitfalls, and upstream changes. Use when starting a new feature, investigating an unfamiliar area, or the user asks to research something.

    Also in 5 projects that use this

  • SentrySkill

    skill-creator

    Create new agent skills following the Agent Skills specification. Use when asked to "create a skill", "add a new skill", "write a skill", "make a skill", "build a skill", or scaffold a new skill with SKILL.md. Guides through requirements, writing, registration, and verification.

    Also in 5 projects that use this

  • MicrosoftSkill

    handoff

    Compact the current conversation into a handoff document for another agent to pick up.

    Also in 3 projects that use this

  • +5 more from google

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 56 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 4

    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:

    Every commit message and every PR title follows Conventional Commits 1.0.0: <type>[(<scope>)][!]: <description>.

    line 12
  2. Very specific:

    No line starts with a word joined to (, such as foo( or foo( , unless that ( closes before any other (

    line 23
  3. Very specific:

    No paragraph starts with <type>: or <type>(<scope>):, which release-please reads as a second commit.

    line 25
  4. Very specific:

    Never push to main. Every change goes through a PR, and the main-protection ruleset blocks direct pushes for everyone.

    line 33

“Common” is measured: kept by at least 6 in 10 of the 56 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

Conventional commits for langchain-litellm

main takes squash merges only. The squash commit's subject is the PR title and its body is the PR description; branch commits never reach main. release-please computes the version and changelog from those squash commits, and label-pr.yml labels the PR from its title. A malformed title means a wrong release, a missing changelog entry and a wrong label.

Titles and commit messages

  1. Every commit message and every PR title follows Conventional Commits 1.0.0: <type>[(<scope>)][!]: <description>.
    • The type is lowercase.
    • The scope is optional, names one subsystem such as chat_models, router, deps or release, and is never an empty ().
    • The description is imperative, starts with a lowercase verb, keeps identifiers and proper nouns in their own case, and has no trailing period.
    • CI checks the type, the header shape and the empty scope. The rest is on the author.
  2. The type is one of the types: in .github/workflows/pr-title-lint.yml, which is the source of truth: check it. Today: feat, fix, docs, style, refactor, perf, test, build, ci, chore, revert.
  3. A breaking change puts ! before the colon, and only a breaking change carries it. label-pr.yml detects only !, so a BREAKING CHANGE: footer alone gets no breaking label.

PR descriptions

  1. The PR description must parse as a commit body. It becomes the squash commit body, whoever merges can rewrap it, and release-please tries every body line as a footer.
    • No line starts with a word joined to (, such as foo( or `foo(, unless that ( closes before any other ( and before the line ends. Code blocks count, and indenting their lines does not help: the squash rewraps the description at about 72 columns and drops the indent.
    • So a call with a space or a nested paren inside, such as foo(a, b), stays out of the description, code blocks included, or the PR gets a BEGIN_COMMIT_OVERRIDE block before the merge. A call like foo(x) is safe anywhere.
    • No paragraph starts with <type>: or <type>(<scope>):, which release-please reads as a second commit. Do not repeat the title as a line of the description.
    • Close an open issue with Fixes #N. Link an issue that is already closed with Follow-up to #N, never Fixes or Refs, because release-please prints every footer reference as "closes #N".
    • release-please drops a commit it cannot parse, and the release-please job still succeeds. The release-notes job then fails on the push to main. The fix is a BEGIN_COMMIT_OVERRIDE block in the PR description, added before the merge when the description is at risk, or afterwards followed by a re-run of Release.

Dependencies and branches

  1. A dependency change that users need released uses fix(deps). build has no entry in changelog-sections in release-please-config.json, so it cuts no release.
  2. Branch names are <type>/<kebab-case-slug> with the same type as the PR title, for example ci/label-dependabot-security-prs. Bot branches are exempt.
  3. Never push to main. Every change goes through a PR, and the main-protection ruleset blocks direct pushes for everyone.

Labels

  1. Respect label ownership.
    • PR type labels feat, fix, docs, refactor, chore, ci, test, perf, build and breaking: label-pr.yml applies them from the title when a PR is opened or edited, and removes any the title no longer implies. Never set them by hand on PRs. style and revert have no label.
    • Bot-owned: autorelease: pending and autorelease: tagged (release-please); dependencies, python, python:uv and github_actions (Dependabot). Never add or remove them.
    • Issue labels: the issue templates apply bug, feature request or task. Everything else is manual triage by a maintainer: duplicate, invalid, wontfix, question, good first issue, help wanted, upstream, security, regression, and type labels such as docs, chore or ci on issues.
    • bug and feature request are for issues, fix and feat for PRs. Never swap them.

Actions

  1. Pin any new or updated GitHub Action to a full commit SHA with a comment naming its full release, such as # v6.1.1, as every existing uses: line is. A pre-commit hook rejects a shorter comment such as # v6.

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.