name: agent-sdk-guide description: Guide for Claude Agent SDK - build custom AI agents powered by Claude. Triggers on agent sdk, claude-agent-sdk, @anthropic-ai/claude-agent, build agent, programmatic agent, embed claude, custom ai agent, agent infrastructure. Covers installation, authentication providers, tool permissions, file-based configuration, and TypeScript/Python code examples.

Claude Agent SDK Guide

Overview

The Claude Agent SDK enables developers to build custom AI agents powered by Claude. It's the same agent infrastructure that powers Claude Code, now available for embedding in your own applications.

Key Value:

Build autonomous agents that perform complex, multi-step tasks
Same battle-tested infrastructure as Claude Code
Production-ready with built-in error handling and monitoring

Installation

TypeScript

npm install @anthropic-ai/claude-agent-sdk

Python

pip install claude-agent-sdk

Core Features

Context Management

Automatic compaction ensures your agent doesn't run out of context during extended operations. The SDK handles token management transparently.

Tool Ecosystem

Built-in capabilities include:

File operations - Read, write, edit files
Code execution - Run bash commands safely
Web search - Access real-time information
MCP extensibility - Add custom tools via Model Context Protocol

Permission Controls

Fine-grained control over agent capabilities:

{
  allowedTools: ['Read', 'Write', 'Bash'],  // Explicit allowlist
  disallowedTools: ['WebSearch'],            // Explicit blocklist
  permissionMode: 'default' | 'strict'       // Overall strategy
}

Production Features

Built-in error handling and retries
Session management for long-running agents
Monitoring and observability hooks
Automatic prompt caching for performance

Authentication Providers

Anthropic API (Default)

export ANTHROPIC_API_KEY=sk-ant-...

Get your key from console.anthropic.com

Amazon Bedrock

export CLAUDE_CODE_USE_BEDROCK=1
export AWS_ACCESS_KEY_ID=...
export AWS_SECRET_ACCESS_KEY=...
export AWS_REGION=us-east-1

Google Vertex AI

export CLAUDE_CODE_USE_VERTEX=1
export GOOGLE_APPLICATION_CREDENTIALS=/path/to/credentials.json
export VERTEX_REGION=us-central1
export VERTEX_PROJECT_ID=your-project

Microsoft Foundry

export CLAUDE_CODE_USE_FOUNDRY=1
# Azure credentials configuration

File-Based Configuration

Agents can leverage Claude Code's file-based architecture for extensibility:

Directory	Purpose
`.claude/agents/`	Subagent definitions (Markdown)
`.claude/skills/`	Skill files (`SKILL.md`)
`.claude/commands/`	Slash commands (Markdown)
`.claude/settings.json`	Hooks configuration
`CLAUDE.md`	Project memory/context

Important: To use project settings, explicitly enable with:

{
  settingSources: ['project']
}

Code Examples

TypeScript: Basic Agent

import { Agent } from '@anthropic-ai/claude-agent-sdk';

const agent = new Agent({
  model: 'claude-sonnet-4-20250514',
  allowedTools: ['Read', 'Write', 'Bash'],
  systemPrompt: `You are a code review assistant.
    Analyze code for bugs, security issues, and best practices.`,
});

const result = await agent.run({
  prompt: 'Review the code in src/auth.ts for security issues',
});

console.log(result.output);

TypeScript: Agent with MCP Tools

import { Agent } from '@anthropic-ai/claude-agent-sdk';

const agent = new Agent({
  model: 'claude-sonnet-4-20250514',
  mcpServers: {
    database: {
      command: 'npx',
      args: ['-y', '@modelcontextprotocol/server-postgres'],
      env: { DATABASE_URL: process.env.DATABASE_URL },
    },
  },
});

const result = await agent.run({
  prompt: 'List all users who signed up in the last 24 hours',
});

Python: Basic Agent

from claude_agent_sdk import Agent

agent = Agent(
    model="claude-sonnet-4-20250514",
    allowed_tools=["Read", "Write", "Bash"],
    system_prompt="""You are a code review assistant.
    Analyze code for bugs, security issues, and best practices.""",
)

result = agent.run(
    prompt="Review the code in src/auth.py for security issues"
)

print(result.output)

Python: Agent with Custom Permissions

from claude_agent_sdk import Agent, PermissionMode

agent = Agent(
    model="claude-sonnet-4-20250514",
    permission_mode=PermissionMode.STRICT,
    allowed_tools=["Read", "Grep", "Glob"],
    disallowed_tools=["Bash"],  # Extra safety
)

result = agent.run(
    prompt="Find all TODO comments in the codebase"
)

Agent Types & Use Cases

Coding Agents

SRE Diagnostics - Analyze logs, identify issues, suggest fixes
Security Auditing - Scan for vulnerabilities, check dependencies
Incident Triage - Correlate alerts, gather context, draft runbooks
Code Review - Enforce standards, catch bugs, suggest improvements

Business Agents

Legal Review - Analyze contracts, flag risks, summarize terms
Financial Analysis - Process reports, identify anomalies, forecast
Customer Support - Handle tickets, route issues, draft responses
Content Creation - Generate docs, edit copy, ensure consistency

Best Practices

Start Simple

Begin with minimal tool access and expand as needed:

// Start here
allowedTools: ['Read']

// Then add as required
allowedTools: ['Read', 'Grep', 'Glob']

// Only if truly needed
allowedTools: ['Read', 'Write', 'Bash']

Use Subagents for Complex Tasks

Break large tasks into specialized subagents:

.claude/agents/
├── code-reviewer.md      # Reviews code changes
├── test-writer.md        # Generates tests
└── doc-generator.md      # Creates documentation

Handle Errors Gracefully

try {
  const result = await agent.run({ prompt });
  if (result.error) {
    console.error('Agent error:', result.error);
    // Implement retry or fallback logic
  }
} catch (error) {
  console.error('SDK error:', error);
  // Handle network/auth issues
}

Monitor Agent Behavior

const agent = new Agent({
  // ... config
  onToolUse: (tool, input) => {
    console.log(`Tool: ${tool}, Input: ${JSON.stringify(input)}`);
  },
  onThinking: (thought) => {
    console.log(`Thinking: ${thought}`);
  },
});

agent-sdk-guide