スキル一覧に戻る

apply-semantic-versioning

pjt222
更新日 Yesterday
2 閲覧
17
2
17
GitHubで表示
メタdesigndata

について

このスキルは、コード変更を分析し、SemVer 2.0.0に従って適切なセマンティックバージョンの更新(メジャー、マイナー、パッチ)を自動的に決定します。破壊的変更の検出、プレリリース識別子、リリース準備のためのビルドメタデータを処理します。変更をマージした後にバージョンの不一致を解決するため、またはリリースにタグを付ける前に使用してください。

クイックインストール

Claude Code

推奨
メイン
npx skills add pjt222/agent-almanac -a claude-code
プラグインコマンド代替
/plugin add https://github.com/pjt222/agent-almanac
Git クローン代替
git clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/apply-semantic-versioning

このコマンドをClaude Codeにコピー&ペーストしてスキルをインストールします

ドキュメント

Apply Semantic Versioning

Determine and apply the correct semantic version bump by analyzing changes since the last release. This skill reads version files, classifies changes as breaking (major), feature (minor), or fix (patch), computes the new version number, and updates the appropriate files. Follows SemVer 2.0.0 specification.

When to Use

  • Preparing a new release and need to determine the correct version number
  • After merging a set of changes and before tagging a release
  • Evaluating whether a change constitutes a breaking change
  • Adding pre-release identifiers (alpha, beta, rc) to a version
  • Resolving disagreement about what version bump is appropriate

Inputs

  • Required: Project root directory containing a version file (DESCRIPTION, package.json, Cargo.toml, pyproject.toml, or VERSION)
  • Required: Git history since the last release (tag or commit)
  • Optional: Commit convention in use (Conventional Commits, free-form)
  • Optional: Pre-release label to apply (alpha, beta, rc)
  • Optional: Previous version if not readable from files

Procedure

Step 1: Read Current Version

Locate and read the version file in the project root.

# R packages
grep "^Version:" DESCRIPTION

# Node.js
grep '"version"' package.json

# Rust
grep '^version' Cargo.toml

# Python
grep 'version' pyproject.toml

# Plain file
cat VERSION

Parse the current version into major.minor.patch components. If the version contains a pre-release suffix (e.g., 1.2.0-beta.1), note it separately.

Got: Current version identified as MAJOR.MINOR.PATCH[-PRERELEASE].

If fail: If no version file is found, check for a VERSION file or git tags (git describe --tags --abbrev=0). If no version exists at all, start at 0.1.0 for initial development or 1.0.0 if the project has a stable public API.

Step 2: Analyze Changes Since Last Release

Retrieve the list of changes since the last tagged release.

# Find the last version tag
git describe --tags --abbrev=0

# List commits since that tag
git log --oneline v1.2.3..HEAD

# If using Conventional Commits, filter by type
git log --oneline v1.2.3..HEAD | grep -E "^[a-f0-9]+ (feat|fix|BREAKING)"

If no tags exist, compare against the initial commit or a known baseline.

Got: A list of commits with messages that can be classified by change type.

If fail: If git history is unavailable or tags are missing, ask the developer to describe the changes manually. Classify based on their description.

Step 3: Classify Changes

Apply the SemVer classification rules:

Change TypeVersion BumpExamples
Breaking (incompatible API change)MAJORRenamed/removed public function, changed return type, removed parameter, changed default behavior
Feature (new backwards-compatible functionality)MINORNew exported function, new parameter with default, new file format support
Fix (backwards-compatible bug fix)PATCHBug fix, documentation correction, performance improvement with same API

Classification rules:

  1. If ANY change is breaking, the bump is MAJOR (resets minor and patch to 0)
  2. If no breaking changes but ANY new features, the bump is MINOR (resets patch to 0)
  3. If only fixes, the bump is PATCH

Special cases:

  • Pre-1.0.0: During initial development (0.x.y), minor bumps may contain breaking changes. Document clearly.
  • Deprecation: Deprecating a function is a MINOR change (it still works). Removing it is MAJOR.
  • Internal changes: Refactoring that does not change the public API is PATCH.

Got: Each change classified as breaking/feature/fix, and the overall bump level determined.

If fail: If changes are ambiguous, err on the side of a higher bump. A conservative major bump is better than a minor bump that breaks downstream code.

Step 4: Compute New Version

Apply the bump to the current version:

CurrentBumpNew Version
1.2.3MAJOR2.0.0
1.2.3MINOR1.3.0
1.2.3PATCH1.2.4
0.9.5MINOR0.10.0
2.0.0-rc.1(release)2.0.0

If a pre-release label is requested:

  • 1.3.0-alpha.1 for first alpha of upcoming 1.3.0
  • 1.3.0-beta.1 for first beta
  • 1.3.0-rc.1 for first release candidate

Pre-release precedence: alpha < beta < rc < (release).

Got: New version number computed following SemVer rules.

If fail: If the current version is malformed or non-SemVer, normalize it first. For example, 1.2 becomes 1.2.0.

Step 5: Update Version Files

Write the new version to the appropriate file(s).

# R: Update DESCRIPTION
# Change "Version: 1.2.3" to "Version: 1.3.0"
// Node.js: Update package.json
// Change "version": "1.2.3" to "version": "1.3.0"
// Also update package-lock.json if present
# Rust: Update Cargo.toml
# Change version = "1.2.3" to version = "1.3.0"

If the project has multiple files that reference the version (e.g., _pkgdown.yml, CITATION, codemeta.json), update all of them.

Got: All version files updated consistently to the new version number.

If fail: If a file update fails, revert all changes to maintain consistency. Never leave version files in a partially updated state.

Step 6: Create Version Tag

After committing the version bump, create a git tag.

# Annotated tag (preferred)
git tag -a v1.3.0 -m "Release v1.3.0"

# Lightweight tag (acceptable)
git tag v1.3.0

Use the project's established tag format:

Got: Git tag created matching the new version.

If fail: If the tag already exists, the version was not properly bumped. Check for duplicate tags with git tag -l "v1.3*" and resolve before proceeding.

Validation

  • Current version was read from the correct version file
  • All commits since the last release were analyzed
  • Each change is classified as breaking, feature, or fix
  • The bump level matches the highest-severity change (breaking > feature > fix)
  • New version follows SemVer 2.0.0 format: MAJOR.MINOR.PATCH[-PRERELEASE][+BUILD]
  • All version files in the project are updated consistently
  • No version was skipped (e.g., 1.2.3 to 1.4.0 without 1.3.0 being released)
  • Git tag matches the new version and project's tag format convention
  • Pre-release suffix, if used, follows correct precedence (alpha < beta < rc)

Pitfalls

  • Skipping minor versions: Going from 1.2.3 directly to 1.4.0 because "we added two features." Each release gets one bump; the number of features does not determine the version.
  • Treating deprecation as breaking: Deprecating a function (adding a warning) is a minor change. Only removing it is a breaking change.
  • Forgetting pre-1.0.0 rules: Before 1.0.0, the API is considered unstable. Some projects bump minor for breaking changes during this phase, but it should be documented.
  • Inconsistent version files: Updating package.json but not package-lock.json, or updating DESCRIPTION but not CITATION. All version references must stay in sync.
  • Build metadata confusion: Build metadata (+build.123) does not affect version precedence. 1.0.0+build.1 and 1.0.0+build.2 have the same precedence.
  • Not tagging releases: Without git tags, future version bumps cannot determine the baseline for change analysis.

Related Skills

  • manage-changelog -- Maintain changelog entries that pair with version bumps
  • plan-release-cycle -- Plan release milestones that determine when version bumps occur
  • release-package-version -- R-specific release workflow that includes version bumping
  • commit-changes -- Commit the version bump with a proper message
  • create-github-release -- Create a GitHub release from the version tag

GitHub リポジトリ

pjt222/agent-almanac
パス: i18n/caveman-lite/skills/apply-semantic-versioning
0
agentsagentskillsai-assisted-developmentclaude-codeskillsteams

関連スキル

content-collections

メタ

このスキルは、Content Collections(Markdown/MDXファイルを型安全なデータコレクションに変換するTypeScriptファーストのツール)の本番環境でテストされた設定を提供します。Zodバリデーションによる型安全性を実現し、ブログ、ドキュメントサイト、コンテンツ重視のVite + Reactアプリケーション構築時にご利用ください。Viteプラグインの設定、MDXコンパイルから、デプロイ最適化、スキーマバリデーションまで、すべてを網羅しています。

スキルを見る

polymarket

メタ

このスキルは、開発者がPolymarket予測市場プラットフォームを活用したアプリケーション構築を可能にします。API統合による取引や市場データの取得に加え、WebSocketを介したリアルタイムデータストリーミングにより、ライブ取引や市場活動を監視できます。取引戦略の実装や、ライブ市場更新を処理するツールの作成にご利用ください。

スキルを見る

creating-opencode-plugins

メタ

このスキルは、開発者がコマンド、ファイル、LSP操作など25種類以上のイベントタイプにフックするOpenCodeプラグインを作成することを支援します。JavaScript/TypeScriptモジュール向けに、プラグイン構造、イベントAPI仕様、および実装パターンを提供します。カスタムイベント駆動ロジックでOpenCode AIアシスタントのライフサイクルをインターセプト、監視、または拡張する必要がある場合にご利用ください。

スキルを見る

sglang

メタ

SGLangは、高性能なLLMサービングフレームワークであり、RadixAttentionプレフィックスキャッシュを活用したJSON、正規表現、エージェントワークフロー向けの高速で構造化された生成を特長とします。特にプレフィックスが繰り返されるタスクにおいて、大幅に高速な推論を実現し、複雑な構造化出力やマルチターン対話に最適です。制約付きデコードが必要な場合や、広範なプレフィックス共有を伴うアプリケーションを構築する場合は、vLLMなどの代替案ではなくSGLangを選択してください。

スキルを見る