Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

发布新版本（Release） · f-is-h | DecimalAI

devopsby f-is-h

发布新版本（Release）

f-is-h/release

Source f-is-h/usage4claude ↗License MIT

发布 Usage4Claude 新版本时使用。当用户说“发布新版本 / 发版 / 出新版 / release / 打 tag 发布 / 准备发版材料”等，用本 skill 引导完成从收集变更、编写 CHANGELOG 与 RELEASE_NOTES、更新版本号、编译验证，到发版 commit、CI 自动发布的完整流程。

SkillScore86Maintained

Automatic

DevOps & SRE pack

Last verified 8/21/2026

What it changes

Passed 19 of 22 test cases with the skill, and 8 of 22 without it.

Without skill36%

With skill86%

Checks it passed that the model alone missed

Recommends targeting the specific test process ID with kill <PID>.
Advises keeping CURRENT_PROJECT_VERSION = $(MARKETING_VERSION) unchanged.
Avoids modifying CURRENT_PROJECT_VERSION or manually fixing the build number.

and 13 more in the full report

11 fixed8 passed either way3 still failing0 broke

Pass-rate lift: +50 pp
Tokens: +168%
Agent turns: 0%
Response time: -34%

Model gemini-3.6-flashMeasured 8/21/2026v119 of 22 cases passed with skill

Tokens, turns and time describe resource use, not a dollar price.

Use this skill

Save it into your project's skills folder. Claude Code picks it up in its next session.

curl -fsSL --create-dirs -o .claude/skills/f-is-h-release/SKILL.md https://app.decimal.ai/s/f-is-h-release/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add f-is-h-release, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"

1 · One-time — install the CLI (docs ↗)

pip install "decimalai>=0.15"

2 · Get the skill — writes .agents/skills/f-is-h-release/ and pins it in .decimal/skills.lock

decimalai skills add f-is-h-release

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/f-is-h-release/SKILL.md https://app.decimal.ai/s/f-is-h-release/SKILL.md

Raw file: /s/f-is-h-release/SKILL.md ↗ · machine-readable summary at /s/f-is-h-release.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 11 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 11

Prompt

When adding a new section ## [3.4.0] - 2025-05-10 to the top of CHANGELOG.md in Usage4Claude, what update must be made at the very bottom of CHANGELOG.md?

What was checked, without and with the skill
Checked	Without skill	With skill
Adds a new version link reference [3.4.0]: https://github.com/f-is-h/Usage4Claude/releases/tag/v3.4.0 at the bottom of CHANGELOG.md.	Missed	Passed

case-15 · gemini-3.6-flash · 8/21/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md209 linessingle file

发布新版本（Release）

Usage4Claude 采用 CI 自动发布：向 main push 一个满足条件的 commit 后， GitHub Actions（.github/workflows/release.yml）自动完成构建、签名、发 Release、更新 Sparkle 更新源。你（Claude）的职责是准备好发版材料并引导用户完成触发， 不代替用户执行发版 commit 与 push。

若根目录存在 NEXT_RELEASE.md，说明本次发版有一次性特殊情况，先读它再按本流程走。

必须先记住的架构事实

两份发布材料，各司其职（详见下一节分工表）：
- CHANGELOG.md — 完整技术档案 + 版本号权威源，不进 Sparkle。
- docs/RELEASE_NOTES.md — 面向用户的发布说明，CI 提取当前版本段落后同时注入 Sparkle 弹窗（appcast.xml 的 <description>）和 GitHub Release 正文。
触发条件：commit message 含 [release]/[RELEASE]，push 到 main，且本次改动包含 CHANGELOG.md 或 docs/RELEASE_NOTES.md（发版通常两者都改）。
Sparkle 弹窗内容来自 docs/RELEASE_NOTES.md（不是 CHANGELOG）。用户在应用内“检查更新” 看到的就是 RELEASE_NOTES 当前版本段落。
docs/RELEASE_NOTES.md 必须有当前版本段落：否则 Sparkle/Release 正文会为空。CI 的 validate 阶段会 grep "^## [X.Y.Z]" fail-fast，但应发版前就写好。
Build 号自动跟随：CURRENT_PROJECT_VERSION = $(MARKETING_VERSION)，Build 恒等于 Version。只改 Version，绝不手动固定 Build，否则 Sparkle 认不出新版本。
appcast.xml 由 CI 维护，绝不手改。
详细背景见 docs/DAILY_RELEASE_WORKFLOW.md、docs/SPARKLE_SETUP.md。

两份材料的分工（关键）

	CHANGELOG.md	docs/RELEASE_NOTES.md
定位	完整技术档案 + 版本号权威源	面向用户的发布说明
收录范围	所有改动，含内部重构、CI、安全加固	只留用户可感知的现象
措辞	可保留技术细节（JWT、actor、base64url 等）	口语化，去技术词
致谢	不加	在相关条目末尾加 `(thanks @author, #N)`
CI 喂给	无（纯档案；validate 从它提版本号）	Sparkle 弹窗 + GitHub Release 正文
何时写	发版前	发版前（不是发布后精修）

规则细节见 docs/CHANGELOG_AND_RELEASE_NOTES_GUIDELINES.md。

流程

Track record

Benchmark

19 / 22

test cases passed, vs 8 / 22 without

Adoption

Users: 0
Times used: 0

Safety and source details

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026

View 15 scanner checks

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

f-is-h / usage4claude

.agents/skills/release/SKILL.md

View on GitHub ↗

License MITImported 8/3/2026Version v1

More skills from the same source and nearby in the registry, with their headline benchmark lift.

Ready to supercharge your AI agents?

See it in action

What's inside

发布新版本（Release）

必须先记住的架构事实

两份材料的分工（关键）

流程

Track record

Benchmark

Adoption

SkillSafety

Source

Related skills