Claude API Skill — Development skill for Claude Code

Claude API Skill (Professional Edition v2.0)

**The most comprehensive guide for building production-grade applications with the Anthropic Claude API.**

Master Messages API, streaming, tool use, vision, extended thinking, batches, prompt caching, optimization, error handling, and architecture patterns.

---

What's Inside

This enhanced skill framework covers everything needed to build scalable AI applications:

Core Concepts

• Model Selection — Comparison matrix and selection guide (Opus, Sonnet, Haiku)
• 7 Key Features — Messages, Streaming, Tool Use, Vision, Extended Thinking, Caching, Batches

Python SDK (Complete Guide)

• Installation & configuration
• Basic messages, streaming, system prompts
• Multi-turn conversations
• Tool use with examples
• Vision (images, multiple images)
• Extended thinking for complex reasoning
• Prompt caching (90% cost savings)
• Batches API (50% cost reduction)

TypeScript SDK (Complete Guide)

• Installation & configuration
• Basic messages, streaming, system prompts
• Tool use with event handlers
• Vision with image loading
• Extended thinking
• Prompt caching
• Batches API

Production Patterns

• Error handling with retry logic
• Token counting before API calls
• Structured JSON output
• Smart caching strategies
• Comprehensive logging and monitoring

Cost Optimization

• Model selection by task ($4.80/M vs $18 vs $90)
• Caching strategy (90% savings)
• Batches API (50% savings)
• Cost examples with real numbers
• Task-specific model selection

Debugging & Monitoring

• Request/response logging
• Token usage tracking
• API metrics collection
• Common pitfalls and solutions
• Quality checklist

---

Quick Start

Python

from anthropic import Anthropic

client = Anthropic()  # reads ANTHROPIC_API_KEY from environment

message = client.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello, Claude!"}]
)
print(message.con