agentleFS
ccoalmccoalm/ccl-skills/skills/release-doc-writer/SKILL.md

release-doc-writer

Skill6 starsChanged 3 months agoApache-2.0View on GitHub

提测文档 / 上线文档 / 发布文档 / test handoff / release doc / launch notes / 补上线范围 / 发布证据 → write, fill, or review test-handoff and release-document substance from Git/config/deploy/verification evidence without owning release orchestration, merge authorization, or production mutation.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill when the task is to write, fill, review, or update test-handoff, release, or launch documentation: delivery scope, config …

    line 8
  2. Distinctive:

    Avoid full-document replacement when comments or anchors exist.

    line 37
  3. Distinctive:

    Do not overwrite unrelated sections.

    line 40
  4. Distinctive:

    If the task includes preparing MR/PR handoff text, the release document should be updated first and the MR description should cite the …

    line 44

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Release Doc Writer(提测 / 上线文档编写)

Use this skill when the task is to write, fill, review, or update test-handoff, release, or launch documentation: delivery scope, config changes, migrations, dependencies, scripts, worker/queue decisions, verification evidence, rollback notes, MR handoff notes, or Feishu/Lark handoff sections.

This skill owns document substance and evidence discipline. It does not authorize deployment, merge, tag, production mutation, branch reset, or overall release orchestration. For full production release coordination, use release-coordination.

Skip / route

  • Full production release lifecycle, MR merge authorization, tag gate, pipeline follow-up, post-release reset → release-coordination.
  • Rollout/canary/rollback strategy or live deploy/config control plane → platform-release-engineering.
  • Risk gate decision before release → feature-risk-router.
  • What tests to run or how to prove behavior → testing-strategy.
  • Active production error or failed smoke → defect-diagnosis.
  • Pure wording polish after facts are fixed → tighten-doc.
  • Actual Feishu/Lark API mechanics → the corresponding document tooling.

Evidence workflow

  1. Establish source range — identify repo/service list, base ref, head ref, and whether the evidence is commit log, changed-file summary, line-level diff, CI output, runtime state, or document read-back.
  2. Write scope from evidence — group by user-visible or operational capability; do not infer scope from branch names, MR titles, memory, or another model's summary alone.
  3. Scope confirmation handoff — if the actual diff/log scope is broader or narrower than the stated release intent, stop and route the release decision back to release-coordination before writing it as settled fact.
  4. Config gate — before writing “no config change”, check config definitions/defaults, usage sites, deployment/config surfaces, and absent-value behavior enough to justify the claim.
  5. Operational sections — inspect touched migrations, jobs, dependencies, external services, workers, queues, scripts, and enable/disable flags; record enabled, intentionally disabled, deferred, and not-run decisions.
  6. Testing scope section — write planned/needed test coverage from confirmed diff; route full matrix or command design to testing-strategy.
  7. Verification section — list only tests, CI, smoke, runtime checks, logs, or metrics actually run or observed. Otherwise say 未验证 / 待验证.
  8. Rollback notes — write concrete rollback constraints evidenced by diff/config/data/runtime state; route strategy design to platform-release-engineering.
  9. Read back edits — after editing collaborative docs, re-fetch the edited section and comment state when available.

Comment-safe document editing

  • Inspect the target section and existing comments before editing.
  • Avoid full-document replacement when comments or anchors exist.
  • Preserve existing structure: table stays table, list stays list, heading hierarchy stays unchanged.
  • Replace placeholders or target cells/blocks narrowly; append only when no placeholder exists.
  • Do not overwrite unrelated sections.

MR handoff boundary

If the task includes preparing MR/PR handoff text, the release document should be updated first and the MR description should cite the document and decisions. Creating or updating an MR is not merge authorization; route merge authorization and lifecycle sequencing to release-coordination.

Minimal checklist

  • Repos/services and base/head refs recorded.
  • Scope based on Git diff/log or equivalent first-hand evidence.
  • Scope mismatch, over-broad branches, or target-ref drift routed back to release-coordination instead of being written as confirmed.
  • Testing scope section output from diff evidence, or imported from release-coordination, and not treated as executed verification.
  • Config/default/usage/deploy surfaces checked.
  • Migrations/jobs/dependencies/scripts/workers/queues checked where touched.
  • Enabled, disabled, deferred, and not-run decisions documented.
  • Verification claims match actual evidence depth.
  • Collaborative doc edits are narrow and comment-safe.
  • Edited section read back.
  • Final response states evidence depth and gaps.

References

  • references/release-evidence-workflow.md
  • references/comment-safe-release-doc.md
  • references/release-testing-scope-section.md

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.