ワンクリックで
api-documentation
">"
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
">"
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | api-documentation |
| description | > |
| license | MIT |
| compatibility | > |
| allowed-tools | Read Write Edit Glob Grep |
| metadata | {"tags":"api-documentation, developer-docs, openapi, swagger, graphql, webhooks, sdk-docs, developer-portal, integration-guides","platforms":"Claude, ChatGPT, Gemini","version":"2.1.0","modernization":"2026-04-13T00:00:00.000Z","structural_hardening":"2026-04-17T00:00:00.000Z"} |
Use this skill when the main job is publishing or refreshing developer-facing API docs that help integrators reach first success, understand reference truth, and stay unblocked as the API evolves.
api-documentation is the documentation-cluster anchor for:
Read these support docs before choosing the mode or output packet:
api-design.technical-writing.user-guide-writing.authentication-setup.changelog-maintenance.api-design instead of inventing docs from vibes.Normalize the request before drafting.
api_documentation_mode:
primary_mode: reference | quickstart | task-guide | sdk-guide | webhook-guide | migration-update
api_style: rest | graphql | webhook | sdk | mixed | unknown
audience: external-developers | internal-integrators | partners | mixed | unknown
source_of_truth: openapi | graphql-schema | code-annotations | tests | mixed | unknown
docs_surface: developer-portal | docs-site | repo-markdown | sdk-site | internal-catalog | unknown
navigation_scope: single-operation | grouped-resource | large-api-surface | portal-section | unknown
maintenance_state: new-docs | refresh | drift-fix | launch-critical
Choose one primary mode per run:
reference → endpoint/schema truth with parameters, fields, errors, limits, and examplesquickstart → fastest path to first successful API call or webhook receipttask-guide → one workflow or integration outcomesdk-guide → language/client-library guidance plus examples and caveatswebhook-guide → delivery, verification, retry, and local-debug guidancemigration-update → changed behavior and transition path on the docs surfaceIf the surface is large, also decide whether the task is about grouping/navigation as much as prose. Do not hide large-surface information architecture inside a generic “write docs” request.
Answer these four questions before writing:
Quick route-out table:
| If the request sounds like... | Use |
|---|---|
| "Design the endpoints / resources / schema before coding" | api-design |
| "Write the architecture doc / ADR / rollout plan / runbook" | technical-writing |
| "Write customer help docs for using the product UI" | user-guide-writing |
| "Explain provider setup / session middleware / token implementation" | authentication-setup |
| "Summarize what shipped in the release" | changelog-maintenance |
| "Write the developer portal / OpenAPI docs / quickstart / webhook guide" | api-documentation |
Do not write API docs from an outdated spec alone. Pull the minimum credible evidence first:
If details are incomplete, label assumptions clearly and produce a docs-gap list instead of faking certainty.
Match the output to the developer job. Do not dump quickstart, reference, SDK docs, and migration notes into one giant page by default.
Common packet choices:
Use references/output-packets-and-navigation.md for mode skeletons and portal/grouping guidance.
Use these rules aggressively:
Before finalizing, record:
Use references/publishing-and-drift-control.md as the anti-drift checklist.
Before shipping, check:
api-design, technical-writing, user-guide-writing, authentication-setup, and changelog-maintenance still explicit?Input: “Write the quickstart for our Orders API so partners can create an API key, send the first POST /orders, and verify the returned order ID.”
Good output shape: chooses quickstart, shows prerequisites and auth early, gives one first-success request plus a success check, and links to the deeper reference surface.
Input: “Our developer portal publishes one huge OpenAPI collection; help us regroup endpoints, expose only partner-safe sections, and add navigation that developers can scan.”
Good output shape: treats this as reference plus navigation/grouping work, proposes grouped sections or selective publishing, preserves truthful reference links, and does not pretend auto-generation alone solves the structure problem.
Input: “Update our webhook docs because invoice.paid now retries for 24 hours, includes attempt_count, and requires HMAC verification on every delivery.”
Good output shape: chooses webhook-guide or migration-update, documents retry and verification behavior, updates affected examples, and keeps implementation details and release-note hygiene out of scope.
Input: “We haven’t decided whether this should be REST or GraphQL, and we need a resource model and pagination plan.”
Good output shape: routes the task to api-design, explains that contract/interface work must happen before docs publication, and may note what docs surfaces will be needed later.
Assist with Colibri: pure-C LLM inference engine for running GLM-5.2 (744B MoE) on consumer machines with ~25 GB RAM. Use when setting up, building, converting models, running inference, configuring expert streaming and caching, optimizing speculative decoding (MTP), GPU integration, and integrating Colibri into production pipelines. Includes build setup, model download & conversion, chat/inference modes, performance tuning, and API integration patterns.
Discover and apply curated prompts from the prompts.chat collection to optimize AI interactions. Use when refining prompt engineering, finding domain-specific prompt templates, improving response quality, or building prompt-based workflows. Triggers on: prompt optimization, prompt templates, prompt engineering, prompt library, curated prompts, prompt discovery, and AI prompt patterns.
Turn ONE topic into a finished Vox-style paper-collage explainer / ad video, end to end on the Atlas Cloud API + local ffmpeg — script, collage keyframes, motion, voice-over, music, captions, all automated. Use this whenever the user wants a "Vox style" video, a paper/torn-paper collage animation, a "motion collage", a narrated explainer or short ad built from AI-generated collage posters, a scrapbook-style tribute, or wants to turn a topic / product / person into a punchy narrated collage video — even if they don't say the word "Vox". Also use when reproducing Stav Zilber / rom1trs / Higgsfield-style collage ad workflows, or when the user asks for a motion collage or a scrapbook-style tribute. Triggers: "vox video", "collage video", "motion collage", "paper collage explainer", "make a collage ad", "turn this topic into a collage video".
Assist with Motion Previs Studio v4: a cross-platform desktop app for AI-film previsualization. Use when setting up, configuring, troubleshooting, or extending motion-previs-studio for pose extraction, depth mapping, camera motion solving, control layer export, and bundle production for AI-video workflows (Seedance, ComfyUI, Blender, Runway, Kling). Includes build setup, feature integration, UI/logic debugging, and export pipeline optimization.
Work with Lapian Notes / 拉片笔记 (github.com/bkingfilm/lapian-notes) — a local- first React/Vite tool that turns a film into an editable shot-by-shot study notebook: local frame extraction, AI-assisted structure analysis (bring your own AI, no API key required), story-line swimlane timeline, structure tree, and audience-emotion curve. Use when the user asks about Lapian Notes, "拉片笔记", "拉片" (shot-by-shot film analysis) tooling, cloning/running this repo (npm run dev, run.bat/run.command), the AI-analysis-package (ZIP) round-trip workflow, or contributing a PR to lapian-notes. Not for generic video editing (use `opencut` for that) or generic film-analysis theory unrelated to this codebase.
Set up, run, and contribute to TokHub (github.com/yaojingang/TokHub) — an open-source AI API relay monitoring, recommendation, and OpenAI-compatible gateway system with L1/L2/L3 channel health probing, usage metering, alerts, audit, and Docker self-hosting. Use when the user asks about TokHub, "AI API 中转站监控", cloning/running the Go + React monorepo (TOKHUB_ROLE, sqlc, TimescaleDB, NATS), the L1/L2/L3 probe algorithm, the OpenAI-compatible `/gateway/v1/*` endpoint, or contributing a PR to TokHub. Do not use for connecting a running agent to a live TokHub instance's own API (that is covered by the project's own bundled `agent-skills/tokhub` skill inside the TokHub repo, not this one).