release-checklist
Skill2 starsChanged 5 months agoMITView on GitHub
发布检查清单——发布前验证、回滚确认。专项技能SKILL,可独立使用。
31 other projects use a version of this.
Adopted by
Most recently taken up by
- TrueImmortal82/open-code-skillstheir own version, first seen Nov 8, 2026
- striderZA/OpenCodeGameStudiostheir own version, first seen Nov 4, 2026
- OneByJorah/VirtOfficetheir own version, first seen Nov 1, 2026
- erikdarlingdata/PerformanceMonitortheir own version, first seen Oct 25, 2026
- juanitourquiza/shipframetheir own version, first seen Oct 23, 2026
- maximhq/bifrosttheir own version, first seen Oct 20, 2026
Used with
Projects with it also use playtest-report, soak-test and 10 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.
CoralGameSkillplaytest-report
Generates a structured playtest report template or analyzes existing playtest notes into a structured format. Use this to standardize playtest feedback collection and analysis.
Also in 12 projects that use this
lilangMaxSkillsoak-test
Generate a soak test protocol for extended play sessions. Defines what to observe, measure, and log during long play sessions to surface slow leaks, fatigue effects, and edge cases that only appear after sustained play. Primarily used in Polish and Release phases.
Also in 9 projects that use this
lilangMaxSkillteam-live-ops
Orchestrate the live-ops team for post-launch content planning: coordinates live-ops-designer, economy-designer, analytics-engineer, community-manager, writer, and narrative-director to design and plan a season, event, or live content update.
Also in 9 projects that use this
CoralGameSkillteam-polish
Orchestrate the polish team: coordinates performance-analyst, technical-artist, sound-designer, and qa-tester to optimize, polish, and harden a feature or area for release quality.
Also in 14 projects that use this
- +4 more from lilangMax
- +4 more from CoralGame
Other versions31 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 31
- Donchitos/Claude-Code-Game-Studios★ 27ksame name
- maximhq/bifrost★ 9.4ksame name
- tsukumijima/KonomiTV★ 1.1ksame name
- erikdarlingdata/PerformanceMonitor★ 526same name
- nkarasiak/qgis-mcp★ 364same name
- pixel-cellar/Claude-Code-Game-Studios★ 324same name
- striderZA/OpenCodeGameStudios★ 118same name
- tranhieutt/software_development_department★ 71same name
- cenconq25/claude-code-app-studio★ 40same name
- OneByJorah/offivo★ 24same name
21 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository26 other files it gives its agents.
- AGENTS.mdAGENTS.md
- code-reviewSkill
- design-reviewSkill
- main-workflowSkill
- okrSkill
- pr-faqSkill
- release-reviewSkill
- requirement-clarificationSkill
- requirement-reviewSkill
- retro-meetingSkill
- retro-reviewSkill
- six-pagerSkill
- software-project-governanceSkill
- stage-architectureSkill
- stage-cicdSkill
- stage-developmentSkill
- stage-infraSkill
- stage-initiationSkill
- stage-maintenanceSkill
- stage-operationsSkill
- stage-releaseSkill
- stage-researchSkill
- stage-selectionSkill
- stage-testingSkill
- tech-reviewSkill
- test-reviewSkill
The file
Files
发布 Checklist
本 skill 用于版本发布前的系统化检查,确保发布内容、回滚方案和验证计划就绪。可独立使用。
发布候选检查与发布完成检查是两个不同的事实边界:候选阶段运行 check-release --version X.Y.Z --require-changelog --lineage-mode candidate,此时 tag 可以尚未创建;发布 commit、tag 和 push 完成后,运行 check-release --version X.Y.Z --require-changelog --lineage-mode released --release-commit <commit>,本地 tag 缺失、tag 指错 commit、remote tag 缺失或不一致均 fail-closed。不得以候选态 PASS 代替发布完成态 lineage 证据。
0.66.0 起还必须验证 declarative release manifest。FEAT/candidate commit 中保持 lifecycle_state: candidate;发布 commit 才可追加唯一 candidate_to_released event 并改为 released。该 commit 必须只有一个 parent,且 parent 是 Git 派生的 candidate commit。release-ledger 的 UNKNOWN 或 BLOCKED 不是通过。Historical backfill 只提供历史信任,不得据此回补 tag;历史 tag 变更必须有独立 DEC。
触发条件
- 版本发布前,需要确认发布就绪状态
- 功能已通过测试(G7 通过),准备进入发布流程
- 用户要求对当前版本做发布前检查
输入
| 输入项 | 说明 | 是否必须 |
| 变更范围 | 本次发布包含的功能和修改 | 必须 |
| 测试报告 | 测试阶段的质量评估结果 | 必须 |
| 发布计划(如有) | 已有的发布时间窗口和步骤 | 推荐 |
执行步骤
第一步:发布范围确认
| # | 检查项 | 通过标准 |
| 1 | 版本号已定义 | 遵循语义化版本(major.minor.patch) |
| 2 | 变更范围已列出 | 所有变更项有清单(commit 列表或 issue 列表) |
| 3 | 变更类型已标注 | 每项变更标注:新功能 / 修复 / 优化 / 破坏性变更 |
| 4 | 发布时间窗口已确定 | 有明确的发布时间计划 |
第二步:变更日志检查
| # | 检查项 | 通过标准 |
| 1 | CHANGELOG 已更新 | 覆盖本次所有变更(commit 列表与 changelog 对照一致) |
| 2 | 破坏性变更已高亮 | 不兼容变更有明确说明和迁移指引 |
| 3 | 依赖变更已记录 | 新增或升级的依赖有版本号和变更原因 |
| 4 | 已知问题已列出 | 已知但未修复的问题有说明和 workaround |
第三步:回滚方案验证
| # | 检查项 | 通过标准 |
| 1 | 回滚方案已编写 | 有具体的回滚步骤(不是"回滚到上一版本"这样的泛化描述) |
| 2 | 回滚方案已验证 | 在测试环境执行过回滚操作,确认可成功 |
| 3 | 数据兼容性 | 回滚后数据是否兼容(数据迁移场景必检) |
| 4 | 回滚影响范围 | 回滚操作的影响已评估,不会造成额外损失 |
参考 Amazon 可逆性分类:
- 可逆发布(灰度/回滚容易):快速推进,回滚方案简化
- 不可逆发布(数据迁移/Schema 变更):深度论证回滚方案,必须有验证记录
第三步补充:Ledger 与 Projection 验证
| # | 检查项 | 通过标准 |
| 1 | Candidate manifest | release-ledger --version X.Y.Z --no-remote PASS,manifest 仍为 candidate |
| 2 | Version projection | SKILL frontmatter 是唯一权威;release-projection PASS |
| 3 | Atomic write evidence | 如执行 release-projection --write,预生成、路径/symlink 校验、原子替换和 rollback 测试通过 |
| 4 | Released transition | release commit 单父且 parent=candidate commit;只有一个 transition event |
| 5 | Tag facts | local/remote tag 均 peel 到 release commit;remote identity 有效 |
| 6 | Historical boundary | historical_backfill 不作为 native PASS;缺 tag 为 missing_requires_decision,无 DEC 不创建 |
第四步:发布后验证计划
| # | 检查项 | 通过标准 |
| 1 | 核心功能验证清单 | 列出发布后需要验证的核心功能项 |
| 2 | 监控指标基线 | 发布前的关键指标已记录(用于对比) |
| 3 | 告警规则就绪 | 发布相关的监控告警已配置 |
| 4 | 验证责任人 | 每项验证有明确的负责人 |
第五步:数据验证计划(A/B 测试与指标对比)
参考 字节跳动数据驱动——每次发布必须有可衡量的影响评估,不接受"看起来不错"。
| # | 检查项 | 通过标准 |
| 1 | 核心指标基线已记录 | 发布前的关键指标(错误率/延迟/活跃度/转化率)有基线数据 |
| 2 | A/B 测试方案已定义(如适用) | 实验组/对照组分流比例、测试时长、统计显著性阈值已明确 |
| 3 | 成功标准已量化 | 发布后期望的指标变化方向和幅度有预设(如"错误率降低 50%"或"不劣化") |
| 4 | 观察期已确定 | 发布后观察窗口(≥30 分钟 for hotfix,≥24 小时 for 功能发布)已定义 |
| 5 | 回滚触发条件已明确 | 什么指标变化会触发回滚?(如"错误率上升 >2x 基线""P99 延迟劣化 >50%") |
无数据时的替代判定标准(功能无法 A/B 测试时):
- 基础设施变更(CI 配置/监控/部署脚本)→ 用"部署后 CI 稳定运行 N 次"替代指标对比
- 安全修复 → 用"漏洞扫描清零"替代指标对比
- 内部工具/文档变更 → 用"目标用户(开发者/新成员)能否在 X 分钟内完成目标操作"替代
- 无可量化用户指标的小项目 → 用"核心功能冒烟测试通过 + 无新 bug 报告 48 小时"替代
- 替代标准底线:至少要有一个可验证的成功定义——"无数据"不是不做验证的借口
第六步:发布决策
| 决策 | 条件 | 后续动作 |
| 可以发布 | 全部检查项通过,数据验证计划就绪 | 执行发布 |
| 有条件发布 | 非关键项有遗留,数据验证计划就绪 | 记录遗留项,发布后 48 小时内关闭 |
| 暂缓发布 | 关键检查项未通过或数据验证计划缺失 | 修复后重新检查 |
输出
| 产出物 | 格式 |
| 发布 checklist 完成状态 | 每项检查的通过/未通过状态 |
| 变更日志 | 本次发布的完整变更记录 |
| 回滚方案 | 具体步骤,已在测试环境验证 |
| 发布后验证清单 | 核心功能验证项 + 责任人 |
| 数据验证计划 | A/B 测试方案(如适用)+ 核心指标基线 vs 目标 + 回滚触发条件 |
| 发布决策 | 可以发布 / 有条件发布 / 暂缓发布 |
质量标准
- 发布范围明确,有版本号和变更清单
- 变更日志覆盖所有变更,破坏性变更已高亮
- 回滚方案有具体步骤且已验证
- 发布后验证清单存在且包含核心功能验证项
- 数据验证计划已定义(核心指标基线+成功标准+回滚触发条件;无数据时有替代标准)
- 发布决策有明确理由
- Declarative ledger 与 projection 检查通过;四态结果未被降级包装
独立使用说明
本 skill 可脱离完整生命周期独立使用。
如果项目已初始化 governance(.governance/ 存在)
- 先执行目标锚定检查(读 plan-tracker → 确认成功标准 → 偏离检查)
- 不需要先通过任何 Gate
- 可用于任何版本发布前的检查(不限于正式发布阶段)
- 热修复/紧急发布也可使用本 checklist,但可跳过非必要检查项并记录跳过原因
- 如果后续接入完整工作流,本 skill 的产出自动成为 G9(发布就绪)的证据
如果项目未初始化 governance(.governance/ 不存在)
- 跳过所有治理前置检查
- 直接执行本 skill 的核心流程(发布前 checklist + 回滚方案验证 + 发布决策)
- 执行完毕后提醒用户:"当前项目未初始化 governance。如需自动证据记录、Gate 检查、风险跟踪,运行
/governance-init。"
子工作流映射
| 本 skill 步骤 | 对应 lifecycle 阶段 | 映射说明 |
| 发布范围确认 | 版本发布 | Gate G9 检查项"发布范围是否明确" |
| 变更日志检查 | 版本发布 | Gate G9 检查项"变更日志是否完整" |
| 回滚方案验证 | 版本发布 | Gate G9 检查项"回滚方案是否已验证" |
| 发布后验证计划 | 版本发布 | Gate G9 检查项"发布后验证是否已定义" |
| 数据验证计划 | 版本发布 | Gate G9 检查项"影响评估是否已定义"——字节 A/B 测试文化 |
| 发布决策 | 版本发布 | 发布决策记录映射到决策日志 |
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
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.

