Install this Claude Code skill.
Name: Claude API Integration
Save to: ~/.claude/skills/claude-api-integration.md

---
---
name: claude-api-integration
description: Set up Claude API / Anthropic SDK — streaming, tool use, prompt caching. Use when user says "integrate Claude API", "use Anthropic SDK", "add AI to my app", or "set up Claude".
---

Help the user integrate the Claude API into their application.

## Instructions

### Step 1: SDK Setup
Install the Anthropic SDK:
```bash
# TypeScript/JavaScript
npm install @anthropic-ai/sdk

# Python
pip install anthropic
```

Set the API key:
```
ANTHROPIC_API_KEY=sk-ant-...
```

### Step 2: Basic Usage
```typescript
import Anthropic from "@anthropic-ai/sdk"

const client = new Anthropic()

const message = await client.messages.create({
  model: "claude-sonnet-4-20250514",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Hello, Claude!" }],
})

console.log(message.content[0].text)
```

### Step 3: Streaming
For real-time responses:
```typescript
const stream = client.messages.stream({
  model: "claude-sonnet-4-20250514",
  max_tokens: 1024,
  messages: [{ role: "user", content: prompt }],
})

for await (const event of stream) {
  if (event.type === "content_block_delta" && event.delta.type === "text_delta") {
    process.stdout.write(event.delta.text)
  }
}
```

### Step 4: System Prompts
```typescript
const message = await client.messages.create({
  model: "claude-sonnet-4-20250514",
  max_tokens: 1024,
  system: "You are a helpful coding assistant. Always respond with code examples.",
  messages: [{ role: "user", content: "How do I sort an array in Python?" }],
})
```

### Step 5: Tool Use
Define tools Claude can call:
```typescript
const tools = [{
  name: "get_weather",
  description: "Get current weather for a location",
  input_schema: {
    type: "object",
    properties: {
      location: { type: "string", description: "City name" },
    },
    required: ["location"],
  },
}]

const message = await client.messages.create({
  model: "claude-sonnet-4-20250514",
  max_tokens: 1024,
  tools,
  messages: [{ role: "user", content: "What's the weather in Denver?" }],
})

// Check if Claude wants to use a tool
if (message.stop_reason === "tool_use") {
  const toolUse = message.content.find(b => b.type === "tool_use")
  // Execute the tool, then send result back
}
```

### Step 6: Prompt Caching
For repeated system prompts or large contexts:
```typescript
const message = await client.messages.create({
  model: "claude-sonnet-4-20250514",
  max_tokens: 1024,
  system: [{
    type: "text",
    text: largeSystemPrompt,
    cache_control: { type: "ephemeral" },
  }],
  messages: [{ role: "user", content: "..." }],
})
```

### Step 7: Error Handling
```typescript
import Anthropic from "@anthropic-ai/sdk"

try {
  const message = await client.messages.create({ ... })
} catch (error) {
  if (error instanceof Anthropic.RateLimitError) {
    // Wait and retry with exponential backoff
  } else if (error instanceof Anthropic.APIError) {
    console.error(`API error: ${error.status} ${error.message}`)
  }
}
```

### Step 8: Best Practices
- Use the latest model (`claude-sonnet-4-20250514` or `claude-opus-4-20250514`)
- Set `max_tokens` appropriately (don't default to max)
- Use streaming for user-facing responses
- Implement retry logic with exponential backoff for rate limits
- Cache system prompts when reusing across requests
- Keep API keys server-side only (never expose in frontend)
- Log token usage for cost monitoring