Skills Plugins MCP Prompt Model 博客 我的中心
開発 #github #ai

release

Automate the pySigma release process: analyze changes, determine version bump, update version, tag, push, wait for test PyPI, generate release notes, and create GitHub release. Use when: releasing pySigma, creating a new version, publishing to PyPI, version bump, cut a release.

DeepseekModel キュレーション済みスキル 品質 優秀 · 90 v1.0.0

取得

https://deepseekmodel.com/api/download.php?id=sigmahq-pysigma-github-skills-release-skill-md&format=skill
ダウンロード .skill 標準形式。system_prompt と model_config を収録し、任意の Agent で利用可能
.skill ファイルの system_prompt フィールドの実際の内容。
name release description Automate the pySigma release process: analyze changes, determine version bump, update version, tag, push, wait for test PyPI, generate release notes, and create GitHub release. Use when: releasing pySigma, creating a new version, publishing to PyPI, version bump, cut a release. argument-hint Optional: specify patch, minor, or major to skip auto-detection pySigma Release Automates the full release lifecycle for SigmaHQ/pySigma. Prerequisites gh CLI must be installed and authenticated ( gh auth status ) Working directory must be the pySigma repository root Local main branch must be up to date with remote No uncommitted changes Procedure Follow these steps sequentially . Each step requires user confirmation before proceeding to the next. Step 1: Pre-flight Checks Run git status to confirm clean working tree and correct branch ( main ). Run git fetch github then git status to confirm local main is up to date. Run gh auth status to confirm GitHub CLI is authenticated. If any check fails, inform the user and stop. Step 2: Determine Current Version and Last Release Read the current version from pyproject.toml (the version field under [project] ). Use mcp_github_get_latest_release (owner: SigmaHQ , repo: pySigma ) to get the latest release tag and its publication date. Store the latest release tag (e.g., v1.3.0 ) and date for later steps. Step 3: Analyze Changes Since Last Release Run git log <last_release_tag>..HEAD --oneline to get all commits since the last release. Use mcp_github_search_pull_requests to find merged PRs since the last release date: query: merged:>=<YYYY-MM-DD> base:main owner: SigmaHQ , repo: pySigma sort: created , order: asc Paginate if needed to get all results. Classify each change using the version classification guide . Determine the highest applicable bump level: Any major change → major bump Else any minor change → minor bump Else → patch bump Step 4: Propose Version Bump Present to the user: Current version : (from pyproject.toml) Proposed bump : patch / minor / major Proposed new version : X.Y.Z Change summary organized by category: Breaking changes (if any) New features (if any) Bug fixes / improvements Other (documentation, CI, refactoring) Ask the user to confirm the proposed version or override with a different bump level. If the user provided a bump level as an argument, use it and ask for confirmation instead. Step 5: Bump Version in pyproject.toml Run in terminal: poetry version <new_version> (use the exact version number, e.g., poetry version 1.4.0 ). Verify the change: poetry version -s should print the new version. Commit the version bump: git add pyproject.toml git commit -m "Bump version to <new_version>" Ask the user for confirmation before pushing. Push the commit: git push github main . Step 6: Create and Push Version Tag Create an annotated tag: git tag -a v<new_version> -m "Release v<new_version>" . Push the tag: git push github v<new_version> . Inform the user: "Tag v<new_version> pushed. This triggers the test PyPI deployment via GitHub Actions." Step 7: Wait for Test PyPI Deployment Monitor the workflow run triggered by the tag push: gh run list --workflow=release.yml --limit=1 Check the status of the latest run: gh run watch <run_id> Alternatively, poll with: gh run view <run_id> --json status,conclusion If the run fails , inform the user with the failure details ( gh run view <run_id> --log-failed ) and stop the release process . The user must fix the issue before continuing. If the run succeeds , inform the user and proceed. Step 8: Generate Release Notes Compose release notes in Markdown with the following structure: ## What's Changed ### Breaking Changes - Description (PR #N) ### New Features - Description (PR #N) ### Bug Fixes - Description (PR #N) ### Other Changes - Description (PR #N) **Full Changelog** : https://github.com/SigmaHQ/pySigma/compare/ < old_tag > ...v < new_version > Rules for generating notes: Use PR titles and descriptions as the primary source for each entry. Link each entry to its PR: (#N) where N is the PR number. Omit empty sections. If a PR addresses a GitHub issue, mention it: Fixes #N . Keep entries concise — one line per change. Present the draft to the user and let them review and amend the release notes before proceeding. Step 9: Create GitHub Release Ask the user for final confirmation before creating the release. Create the release: gh release create v<new_version> --title "v<new_version>" --notes "<release_notes>" If notes are long, write them to a temp file and use: gh release create v<new_version> --title "v<new_version>" --notes-file <tempfile> Inform the user: "GitHub release created. This triggers the production PyPI deployment." Provide the release URL: https://github.com/SigmaHQ/pySigma/releases/tag/v<new_version> Step 10: Post-Release Verification Check the production PyPI workflow: gh run list --workflow=release.yml --limit=1 Once completed, verify the package is available: pip index versions pySigma Or direct the user to: https://pypi.org/project/pySigma/<new_version>/ Error Recovery Version bump pushed but tag failed : Create and push the tag manually. Test PyPI failed : Fix the issue, delete the tag ( git push --delete github v<new_version> && git tag -d v<new_version> ), amend or re-commit, and restart from Step 6. Release creation failed : Retry gh release create or create manually at https://github.com/SigmaHQ/pySigma/releases/new . Reference Version Classification Guide — Rules for determining version bump level. Release workflow — GitHub Actions workflow that publishes to PyPI.
このスキルを起動するキーワード。クリックでコピーできます。

このスキルにはトリガーワードがありません。

ダウンロードした .skill に含まれるフィールド。
フィールド 説明
formatフォーマット識別子(skill/v1)
skill_idスキル固有 ID
nameスキル名
versionバージョン
description説明
categoryカテゴリ(配列)
trigger_wordsトリガーワード
tagsタグ
sourceソース
source_urlソース URL(本ページ)
exported_atエクスポート日時(ダウンロード毎)
system_promptシステムプロンプト本文
model_configモデル設定:provider / model / temperature / max_tokens / top_p
examplesサンプル
install_guide各プラットフォームの導入説明(Coze / Dify / Claude / カスタム)
同じスキルを各プラットフォーム形式で出力できます。
.skill 標準形式。system_prompt と model_config を収録し、任意の Agent で利用可能 ダウンロード
.skillpro 拡張形式。scripts / tools / dependencies / hooks を含む ダウンロード
.json 純粋な JSON 出力。system_prompt とモデル設定のみ ダウンロード
Coze frontmatter 付き Markdown。Coze へのインポート用 ダウンロード
Dify Dify DSL。アプリ作成後にそのままインポート ダウンロード

每日精选 Skill 推荐,免费送到你邮箱

输入邮箱,每天接收一个精选 AI Agent 技能推荐。完全免费,持续更新。

提交后我们会发送一封确认邮件,点击邮件里的链接才会开始收信。

完全免费,取消任意时间。我们不会发送垃圾邮件。