菜单
GitHub release creation for cc-foundry plugins: per-plugin tags, release notes as the changelog, and post-push publishing. Invoke whenever task involves pushing commits to master, releasing or publishing a plugin version, bumping plugin versions, or writing release notes.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/xobotyi/cc-foundry --skill release命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
基于 SOC 职业分类
正在显示 SKILL.md
| name | release |
| description | GitHub release creation for cc-foundry plugins: per-plugin tags, release notes as the changelog, and post-push publishing. Invoke whenever task involves pushing commits to master, releasing or publishing a plugin version, bumping plugin versions, or writing release notes. |
Every plugin version bump that lands on master ships with a GitHub release. The release notes are the plugin's changelog
— there are no CHANGELOG.md files (see docs/adr/0005). A pushed version bump without a release is an incomplete ship.
.dev/ tooling, this skill) get no release.git diff <old-origin-head>..origin/master -- '*/plugin.json'. When the old
head is unknown, compare each plugin's plugin.json version against its latest <plugin>-v* tag from
gh release list.plugin.json version — the release target<plugin>-v*, if any — the notes range and compare-link anchorgit log <prev-tag>..<target> -- plugins/<name>/ plus the diff of that range. For a first
release, use the commits of the bump itself.<plugin>-v<X.Y.Z>, e.g. the-coder-v1.5.0<plugin> v<X.Y.Z>, e.g. the-coder v1.5.0git tag — gh release create creates the tag on the remote at publish time.Written for people who install the plugin, not for repo contributors.
### section per component: ### `<name>` skill, ### `<name>` output style —
same pattern for hooks, commands, and agents. A plugin bumps its version even when a single component moved; the
segments tell the installer exactly which parts did.### Plugin section.**Added:**, **Changed:**, **Fixed:**,
**Removed:** — ordered Added → Changed → Fixed → Removed.**Full diff:** https://github.com/xobotyi/cc-foundry/compare/<prev-tag>...<new-tag>Sharpens commit-message validation and adds trailer support.
### `commit-message` skill
- **Added:** `Task:` and `Refs:` trailer validation — malformed trailers now fail validation instead of landing
silently.
- **Changed:** subject-length limit aligned with git conventions: 50 soft / 72 hard (was a flat 72).
### `commit` skill
- **Fixed:** validator crashed on commit messages containing unescaped backticks.
**Full diff:** https://github.com/xobotyi/cc-foundry/compare/git-commit-v1.1.3...git-commit-v1.1.4
Write the notes to a scratchpad file first — --notes-file avoids shell-quoting damage:
gh release create <plugin>-v<X.Y.Z> \
--target <bump-commit-sha> \
--title "<plugin> v<X.Y.Z>" \
--notes-file <path-to-notes.md>
--draft (visible to collaborators only; the tag is created at publish, not
at draft creation). Publish with gh release edit <tag> --draft=false — re-set --target to the bump commit first if
the draft was created before the push.gh release view <tag> — confirm title, target, and notes.plugin.json version equals the latest tag.Design Claude Code output styles: role, voice, and behavioral rules injected into the system prompt. Invoke whenever task involves any interaction with output styles — creating, editing, evaluating, or changing how Claude communicates.
CLAUDE.md instruction quality: writing effective project instructions, diagnosing why Claude ignores rules, routing content to the right layer, progressive disclosure, and systematic improvement. Invoke whenever task involves any interaction with CLAUDE.md files — creating, writing, reviewing, auditing, improving, optimizing, or debugging instruction compliance.
Prompt design techniques for LLMs: structure, examples, reasoning patterns, and optimization. Invoke whenever task involves any interaction with AI instructions — crafting, debugging, improving, or evaluating prompts for skills, agents, output styles, or system configurations.