agentleFS
X-T-E-RX-T-E-R/kiki/.agents/skills/gen-docs/SKILL.md

gen-docs

Skill53 starsChanged 14 days agoMITView on GitHub

Update mirrored Kimi Code CLI user guides, configuration, and reference pages. Use when a coherent implementation candidate has confirmed user impact and selected exact non-changelog pages for catch-up. Do not use for release changelog synchronization.

  • Reads credentials

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 7 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Update the user-facing behavior views already selected by the documentation semantic owner.

    line 8
  2. Distinctive:

    Stop and report the missing input when any of these is absent:

    line 14
  3. Distinctive:

    Fix scope and evidence: Read the candidate diff, owning implementation, claim-matched tests, and the selected existing sections.

    line 24
  4. Distinctive:

    Update the selected source view: Follow the surrounding page structure and the terminology, typography, example-data, and Public-Facing …

    line 26
  5. Very specific:

    Produce and approve the mirror: Use translate-docs to write the corresponding docs/en/ or docs/zh/ view.

    line 27
  6. Distinctive:

    Verify: Read both locale sections, compare their reader-relevant claims, and run:

    line 28
  7. Very specific:

    Keep non-changelog pages mirrored under docs/en/ and docs/zh/.

    line 40
  8. Distinctive:

    Update only pages and sections selected by the accepted semantic boundary.

    line 41
  9. Very specific:

    Use neutral public examples such as https://api.example.com/v1, example.test, and YOUR_API_KEY.

    line 42
  10. Very specific:

    Do not edit either changelog page.

    line 43
  11. Distinctive:

    Do not claim that structural checks prove prose accuracy, completeness, lineage, or translation quality.

    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

Gen Docs

Update the user-facing behavior views already selected by the documentation semantic owner. This workflow begins after implementation and claim-matched tests establish the behavior; it does not block early coding.

Release changelog curation is a separate post-release workflow owned by sync-changelog. Do not read or edit either locale's release-notes/changelog.md through this skill.

Prerequisites

Stop and report the missing input when any of these is absent:

  • a coherent candidate boundary and user-impact record from the documentation lifecycle;
  • owning implementation plus a claim-matched test for every behavior to document;
  • exact affected page paths or an accepted semantic packet that selects them;
  • docs/en/, docs/zh/, docs/.vitepress/config.ts, and the documentation directory's governing instructions;
  • the translate-docs skill for meaning-bearing bilingual synchronization.

Workflow

  1. Fix scope and evidence. Read the candidate diff, owning implementation, claim-matched tests, and the selected existing sections. Do not infer behavior from commit messages, PR titles, changesets, or changelog prose.
  2. Confirm reader decisions. For each selected page, state what the reader needs to act, choose, recover, or verify: supported behavior, prerequisites, defaults, configuration, compatibility, migration, observable limits, and failure recovery where applicable. Skip internal refactors, tests, CI, type-only changes, and build changes with no user decision.
  3. Update the selected source view. Follow the surrounding page structure and the terminology, typography, example-data, and Public-Facing Content rules owned by the documentation instructions. Weave changes into existing sections instead of creating a feature diary.
  4. Produce and approve the mirror. Use translate-docs to write the corresponding docs/en/ or docs/zh/ view. Preserve paths, heading levels, lists, code blocks, identifiers, examples, and link shapes. Translation is meaning-bearing work and must be reviewed before any frozen mechanical placement handoff.
  5. Verify. Read both locale sections, compare their reader-relevant claims, and run:

    pnpm docs:check-governance
    pnpm -C docs run build

    Inspect the bounded docs diff. The governance checker proves structure only; the claim-matched tests and semantic review establish behavioral accuracy, while bilingual review establishes translation quality.

  6. Record completion. Mark the selected user views completed, or keep each remaining user view classified as user with an accountable owner and a target no later than default-on or public release.

Rules

  • Keep non-changelog pages mirrored under docs/en/ and docs/zh/.
  • Update only pages and sections selected by the accepted semantic boundary.
  • Use neutral public examples such as https://api.example.com/v1, example.test, and YOUR_API_KEY; do not expose internal endpoints, keys, accounts, or service names.
  • Do not edit either changelog page. After a release succeeds, route changelog work to sync-changelog, whose upstream source is apps/kimi-code/CHANGELOG.md.
  • Do not claim that structural checks prove prose accuracy, completeness, lineage, or translation quality.

Stop conditions

  • The candidate, affected user view, behavior evidence, or claim-matched test is missing.
  • Source and tests disagree, or the selected page would contradict another canonical view.
  • A new page, navigation target, claim, or term requires a semantic decision that the accepted packet did not make.
  • The requested work is changelog synchronization or would cross the supplied write scope.

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.