Skip to main content
Alison Aquinas logoAlison's LLM Skills Marketplace

claude-docs

Included in skill bundleshared-skillsView on GitHub ↗

Files

SKILL.mdagentsreferences

Install

Install only this skill with npx skills
npx skills add alisonaquinas/llm-shared-skills --skill 'claude-docs' -g -y
Install the containing skill bundle
/plugin install shared-skills@llm-skills
Download claude-docs-skill.zip
This skill is bundled inside shared-skills. Use npx skills when you only want this skill, or install the bundle once to make every included skill available through the plugin marketplace flow. Browse the full skill bundle repository at github.com/alisonaquinas/llm-shared-skills.

Invoke

Invoke this skill after installation
/shared-skills:claude-docs

SKILL.md

name: claude-docs description: > Use this skill to look up, navigate, or find anything in the official Anthropic Claude API documentation at platform.claude.com/docs. Trigger when the user asks about Claude API endpoints, model IDs, pricing, prompt engineering, tool use, extended thinking, adaptive thinking, context windows (200K, 1M context), prompt caching (TTL, workspace-level isolation), vision, files API, batch API, the Agent SDK, rate limits, authentication, error codes, or any Anthropic/Claude developer reference. Also trigger for questions like "where is the Claude doc for X", "what does the Claude API reference say about Y", "find the Anthropic docs on Z", or any request to look something up in Claude/Anthropic documentation.

Claude API Docs

Reference skill for navigating official Anthropic Claude documentation at https://platform.claude.com/docs/en/

Full page index: https://platform.claude.com/docs/llms.txt

Intent Router

TopicReference fileLoad when...
Getting started, intro, quickstartreferences/getting-started.mduser needs first API call, setup, or intro overview
Models, model IDs, pricingreferences/models.mduser asks about model names, context windows, pricing, or which model to use
Prompt engineering, tool use, visionreferences/build-with-claude.mduser asks about prompting, function/tool use, image input, streaming, context
Extended thinking, adaptive thinkingreferences/extended-thinking.mduser asks about thinking tokens, budget, extended/adaptive thinking
Agent SDK, sub-agents, orchestrationreferences/agent-sdk.mduser asks about building agents, multi-agent, orchestration, Agent SDK
REST API endpoints, auth, rate limitsreferences/api-reference.mduser asks about HTTP endpoints, authentication, rate limits, errors, SDKs

Key URLs at a Glance

Root               https://platform.claude.com/docs/en/
Intro              https://platform.claude.com/docs/en/intro
Get started        https://platform.claude.com/docs/en/get-started
Models             https://platform.claude.com/docs/en/about-claude/models/overview
Pricing            https://platform.claude.com/docs/en/about-claude/pricing
Prompt engineering https://platform.claude.com/docs/en/build-with-claude/prompt-engineering
Tool use           https://platform.claude.com/docs/en/build-with-claude/tool-use
Extended thinking  https://platform.claude.com/docs/en/build-with-claude/extended-thinking
Context windows    https://platform.claude.com/docs/en/build-with-claude/context-windows
Vision             https://platform.claude.com/docs/en/build-with-claude/vision
Agent SDK          https://platform.claude.com/docs/en/agent-sdk/overview
API reference      https://platform.claude.com/docs/en/api/

Current Model IDs (quick reference)

Authoritative list, including AWS Bedrock and GCP Vertex IDs and current GA status: https://platform.claude.com/docs/en/about-claude/models/overview

Use the canonical overview page when picking a model or constructing a request — it reflects the current GA generation without going stale here. For the current deprecation schedule, see https://platform.claude.com/docs/en/about-claude/model-deprecations.

For a curated cross-platform map (Bedrock + Vertex IDs, capability flags, selection guidance) see references/models.md.

Quick Start

The lookup workflow for almost any Claude API question is the same: pick the matching reference file from the Intent Router, then WebFetch the canonical URL. Two starter recipes:

Minimal first request (Python SDK)

import anthropic

client = anthropic.Anthropic()  # reads ANTHROPIC_API_KEY from env

message = client.messages.create(
    model="<current-sonnet-id>",  # pick from /about-claude/models/overview
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello, Claude"}],
)
print(message.content)

Minimal first request (REST / curl)

curl https://api.anthropic.com/v1/messages \
  --header "x-api-key: $ANTHROPIC_API_KEY" \
  --header "anthropic-version: 2023-06-01" \
  --header "content-type: application/json" \
  --data '{
    "model": "<current-sonnet-id>",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Hello"}]
  }'

For anything beyond a first call (tool use, prompt caching, extended / adaptive thinking, batch processing, files API, Agent SDK), load the matching reference file from the Intent Router and WebFetch the linked page on platform.claude.com.

← Back to marketplace