Use this skill when you need documentation for a third-party library, SDK, or API before writing code that uses it — for example, "use the OpenAI API", "call the Stripe API", "use the Anthropic SDK", "query Pinecone", or any time the user asks you to write code against an external service and you need current API reference. Fetch the docs with chub before answering, rather than relying on training knowledge.
From claude-code-handbooknpx claudepluginhub lioartoil/claude-code-handbookThis skill uses the workspace's default tool permissions.
Dispatches parallel agents to independently tackle 2+ tasks like separate test failures or subsystems without shared state or dependencies.
Executes pre-written implementation plans: critically reviews, follows bite-sized steps exactly, runs verifications, tracks progress with checkpoints, uses git worktrees, stops on blockers.
Guides idea refinement into designs: explores context, asks questions one-by-one, proposes approaches, presents sections for approval, writes/review specs before coding.
When you need documentation for a library or API, fetch it with the chub CLI
rather than guessing from training data. This gives you the current, correct API.
chub search "<library name>" --json
Pick the best-matching id from the results (e.g. openai/chat, anthropic/sdk,
stripe/api). If nothing matches, try a broader term.
chub get <id> --lang py # or --lang js, --lang ts
Omit --lang if the doc has only one language variant — it will be auto-selected.
Read the fetched content and use it to write accurate code or answer the question. Do not rely on memorized API shapes — use what the docs say.
After completing the task, if you discovered something not in the doc — a gotcha, workaround, version quirk, or project-specific detail — save it so future sessions start smarter:
chub annotate <id> "Webhook verification requires raw body — do not parse before verifying"
Annotations are local, persist across sessions, and appear automatically on future
chub get calls. Keep notes concise and actionable. Don't repeat what's already in
the doc.
Always rate the doc after using it. This helps authors fix outdated or incorrect docs and prioritize improvements. Include a label and a brief comment explaining what was good or what needs fixing — specific feedback is the most valuable.
chub feedback <id> up --label accurate "Clear examples, models are current"
chub feedback <id> down --label outdated "Lists gpt-4o as latest but gpt-5.4 is out"
Available labels: outdated, inaccurate, incomplete, wrong-examples,
wrong-version, poorly-structured, accurate, well-structured, helpful,
good-examples.
If you notice the doc has wrong model names, deprecated APIs, missing features, or incorrect code patterns, always leave a downvote with details so it can be fixed.
| Goal | Command |
|---|---|
| List everything | chub search |
| Find a doc | chub search "stripe" |
| Exact id detail | chub search stripe/api |
| Fetch Python docs | chub get stripe/api --lang py |
| Fetch JS docs | chub get openai/chat --lang js |
| Save to file | chub get anthropic/sdk --lang py -o docs.md |
| Fetch multiple | chub get openai/chat stripe/api --lang py |
| Save a note | chub annotate stripe/api "needs raw body" |
| List notes | chub annotate --list |
| Rate a doc | chub feedback stripe/api up |
chub search with no query lists everything available<author>/<name> — confirm the ID from search before fetching--lang, chub will tell you which are available