Anthropic Cookbook

Provides production-ready Jupyter notebooks (.ipynb files) that demonstrate Claude API capabilities through runnable code examples. Each notebook is structured as a self-contained, copy-paste-ready implementation pattern for specific features like tool use, RAG, or multimodal processing. The notebooks serve as both documentation and functional code templates that developers can immediately adapt to their own projects.

Implements a YAML-based registry (registry.yaml) that catalogs all cookbook notebooks with structured metadata including category, tags, author, and description. This enables programmatic discovery, automated validation workflows, and machine-readable capability mapping without requiring manual documentation updates. The registry acts as a single source of truth for content organization and enables tooling to validate notebook compliance.

Implements automated validation infrastructure (scripts/validate_notebooks.py) that ensures all cookbook notebooks remain functional and compliant with standards. Validation checks include notebook structure, API usage correctness, metadata consistency, and execution tests. Integrates with CI/CD pipeline to catch breaking changes and maintain quality across the cookbook collection.

Provides structured contribution guidelines and tooling for adding new notebooks to the cookbook. Includes Claude Code slash commands (.claude/commands/add-registry.md) that semi-automate registry entry creation, GitHub pull request templates that enforce metadata requirements, and contributor documentation (CONTRIBUTING.md). Enables consistent, high-quality contributions without manual registry editing.

Demonstrates advanced RAG patterns using LlamaIndex as an abstraction layer over vector databases and retrieval strategies. Notebooks show how to implement hybrid search (combining keyword and semantic search), multi-hop retrieval (chaining multiple retrieval steps), reranking, and query expansion. Covers integration with multiple vector databases (Pinecone, Weaviate, Chroma) without rewriting core logic.

Provides examples for processing audio and voice input with Claude, including audio transcription, voice analysis, and audio-to-text workflows. Notebooks demonstrate how to encode audio files, send them to Claude, and extract structured information from audio content. Covers use cases like meeting transcription, voice command processing, and audio content analysis.

Provides executable examples demonstrating Claude's tool-calling capability through function schema definitions, parameter binding, and multi-turn interaction patterns. Notebooks show how to define tool schemas (JSON Schema format), handle tool calls in API responses, execute tools, and feed results back to Claude for iterative problem-solving. Covers both simple single-tool scenarios and complex multi-tool orchestration patterns.

Provides end-to-end RAG implementation patterns including document ingestion, vector embedding, semantic search, and context injection into Claude prompts. Notebooks demonstrate integration with vector databases (Pinecone, Weaviate, etc.) via LlamaIndex abstraction layer, showing how to build retrieval systems that augment Claude's knowledge with external documents. Covers both basic RAG (simple retrieval + prompt injection) and advanced patterns (hybrid search, reranking, multi-hop retrieval).

Provides examples demonstrating Claude's prompt caching feature, which reduces latency and costs for repeated requests with large static context. Notebooks show how to structure prompts to maximize cache hits (placing cacheable content in system prompts or message prefixes), measure cache effectiveness, and calculate cost savings. Covers use cases like multi-turn conversations with large system prompts, batch processing with shared context, and document analysis workflows.

Provides examples for Claude's vision capabilities, including image analysis, OCR, visual question answering, and image-based document processing. Notebooks demonstrate how to encode images (base64 or URLs), construct vision prompts with image context, and extract structured information from visual content. Covers use cases like document scanning, diagram interpretation, and visual data extraction.

Provides examples for Anthropic's Batch API, which enables asynchronous, cost-reduced processing of multiple requests. Notebooks demonstrate how to structure batch requests, submit them for processing, poll for completion, and retrieve results. Shows cost savings calculations and use cases like bulk document analysis, large-scale content generation, and overnight processing jobs.

Provides examples for implementing content moderation using Claude, including classification of harmful content, safety filtering, and policy enforcement. Notebooks demonstrate how to use Claude as a moderation classifier, structure moderation prompts, and integrate moderation into application workflows. Covers use cases like user-generated content filtering, prompt injection detection, and safety policy enforcement.

Provides examples for generating embeddings using Anthropic's embedding models and implementing semantic search. Notebooks demonstrate how to embed documents and queries, compute similarity scores, and build search systems. Shows integration with vector databases and includes patterns for chunking documents, managing embedding costs, and handling large-scale embedding workloads.

Provides patterns for managing conversation history and context in multi-turn interactions with Claude. Notebooks demonstrate how to structure message arrays, manage token budgets across turns, implement conversation memory, and handle context window limits. Covers use cases like chatbots, interactive agents, and stateful dialogue systems.