본문으로 건너뛰기
cccskills
무료GitHub에 공개

docs-sync

Audit or update authored English SDK documentation within a requested topic or diff using implementation evidence.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

연동 도구OpenAI
설치 방법 보기

포함된 파일(3)

  • SKILL.md5.7 KB
  • agents/openai.yaml232 B
  • references/doc-coverage-checklist.md2.8 KB

SKILL.md(원문)

설치하기 전에 에이전트에게 전달되는 지시 내용을 확인할 수 있습니다.

Docs Sync

Overview

Identify documentation gaps and inaccuracies within the requested topic or diff using implementation evidence.

Scope and authorization

For audit or proposal-only requests, report findings without editing. When the user requests updates or has approved a plan, complete the scoped edits and verification without asking again. Ask only for unresolved behavior, scope, release timing, or additional authority. Apply the repository's Documentation Release Timing policy before including unreleased behavior in docs/.

Workflow

  1. Confirm scope and base branch

    • Identify the current branch and default branch (usually main).
    • Prefer analyzing the current branch to keep work aligned with in-flight changes.
    • Use a branch diff only for a branch-scoped request. A requested topic or released-doc correction remains in scope even when unrelated to the current branch diff.
    • Avoid switching branches if it would disrupt local changes. Prefer read-only inspection such as git show main:<path>. If a separate checkout is genuinely required, stop and obtain the explicit approval required by AGENTS.md before creating or switching a worktree.
  2. Build a feature inventory from the selected scope

    • Bound the inventory to the requested topic or diff. Inventory the full surface only for an explicitly comprehensive audit.
    • For branch-scoped work, inspect additions, changes, and removals relative to the intended base.
    • Focus on user-facing behavior: public exports, configuration options, environment variables, CLI commands, default values, and documented runtime behaviors.
    • Capture evidence for each item (file path + symbol/setting).
    • Use targeted search to find option types and feature flags (for example: rg "Options", rg "process.env", rg "export").
    • When the topic involves OpenAI platform features, invoke $openai-knowledge to pull current details from the OpenAI Developer Docs MCP server instead of guessing, while treating the SDK source code as the source of truth when discrepancies appear.
    • For MCP SDK (modelcontextprotocol/typescript-sdk) or Vercel AI SDK (@ai-sdk/*) topics, optionally use Deepwiki MCP for quick lookups, and still treat the SDK source code as the source of truth.
  3. Doc-first pass: review existing pages

    • Walk each relevant page under docs/src/content/docs (excluding docs/src/content/docs/openai).
    • Identify missing mentions of important, supported options (opt-in flags, env vars), customization points, or new features from packages/.
    • Propose additions where users would reasonably expect to find them on that page.
  4. Code-first pass: map features to docs

    • Review the current docs information architecture under docs/src/content/docs.
    • Determine the best page/section for each feature based on existing patterns and package boundaries.
    • Identify features that lack any doc page or have a page but no corresponding content.
    • Note when a structural adjustment would improve discoverability.
  5. Detect gaps and inaccuracies

    • Missing: features/configs present in main but absent in docs.
    • Incorrect/outdated: names, defaults, or behaviors that diverge from main.
    • Structural issues (optional): pages overloaded, missing overviews, or mis-grouped topics.
  6. Classify the proposed complete diff

    • Use the Documentation Change Verification Tiers in AGENTS.md as the single source of truth.
    • Classify the complete proposed diff by its highest applicable Editorial, Content, or Structural tier before editing.
    • Keep the tier-specific verification separate from the existing eligibility rules for $implementation-final-review, $code-change-verification, $changeset-validation, and $pr-draft-summary.
  7. Report findings or continue authorized updates

    • For audit-only work, provide evidence, suggested locations, proposed edits, and required verification, then stop.
    • For an update request, complete the authorized edits using the findings and selected verification tier.
  8. Apply authorized changes (English only)

    • Edit only English docs in docs/src/content/docs/**.
    • Exclude docs/src/content/docs/openai from review and updates.
    • Do not edit docs/src/content/docs/ja, docs/src/content/docs/ko, or docs/src/content/docs/zh.
    • Keep changes aligned with the existing docs style and navigation.
    • Do not add TypeScript or TSX fenced blocks directly to MDX. Place every TypeScript or TSX snippet in a compilable source file under examples/docs/<doc-area>/, import it into MDX with ?raw, and render that imported source using the existing docs pattern.
    • If code is not useful enough to maintain as a complete example, explain the behavior in prose instead of adding an unverified snippet.
    • Run the focused verification required by the highest applicable Documentation Change Verification Tier in AGENTS.md. When the complete diff changes a rendered TypeScript or TSX snippet source or its MDX import/rendering, run pnpm -F docs-code build-check and fix issues before handoff.

Output format

Use this template when reporting findings:

Docs Sync Report

  • Doc-first findings
    • Page + missing content → evidence + suggested insertion point
  • Code-first gaps
    • Feature + evidence → suggested doc page/section (or missing page)
  • Incorrect or outdated docs
    • Doc file + issue + correct info + evidence
  • Structural suggestions (optional)
    • Proposed change + rationale
  • Proposed edits
    • Doc file → concise change summary
  • Unresolved decisions, only when needed

References

  • references/doc-coverage-checklist.md

리뷰

아직 리뷰가 없습니다. 사용해 본 소감을 남겨 주세요.

댓글

로그인 없이 쓸 수 있습니다. 사용 후기, 질문, 작동하지 않은 환경 등을 편하게 남겨 주세요.

아직 댓글이 없습니다. 첫 댓글을 남겨 보세요.

보내면 바로 공개됩니다. 이메일 주소나 전화번호 같은 개인 정보는 쓰지 마세요. URL은 2개까지입니다.

0 / 1,000자

이 양식은 reCAPTCHA로 보호되며 Google 개인정보처리방침 및 서비스 약관이 적용됩니다.

같은 저장소의 스킬

Validate changesets in openai-agents-js using LLM judgment against git diffs (including uncommitted local changes). Use when packages/ or .changeset/ are modified, or when verifying PR changeset compliance and bump level.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

openai/openai-agents-js3.9천2026년 10월 16일 업데이트

Run the required final install, build, type, lint, test, and format checks after eligible SDK changes pass review.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

openai/openai-agents-js3.9천2026년 10월 16일 업데이트

Fix the tiny credit-note formatting bug and rerun the exact targeted test command.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

openai/openai-agents-js3.9천2026년 10월 16일 업데이트

Analyze CSV files in /mnt/data and return concise numeric summaries.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

openai/openai-agents-js3.9천2026년 10월 16일 업데이트

Analyze logs and source from a completed repository example run. Never execute or control examples.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

openai/openai-agents-js3.9천2026년 10월 16일 업데이트

Assess a JS SDK release candidate or release plan against the previous release and recommend ship or block.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

openai/openai-agents-js3.9천2026년 10월 16일 업데이트

openai의 스킬 모두 보기

이 스킬의 문제 신고