Claude Agent SDK: Build Autonomous AI Agents

About this course

This course contains the use of artificial intelligence

Build autonomous, production-ready AI agents with the Claude Agent SDK using TypeScript, modern development workflows, and practical real-world projects.

The Claude Agent SDK brings the capabilities behind Claude Code into your own applications, allowing you to build agents that can reason about tasks, inspect files, execute commands, use tools, modify code, maintain sessions, connect to external systems, and work autonomously toward a goal.

In this course, you will move beyond simple prompt-and-response applications and learn how to design agentic systems that plan, act, observe results, and continue working until a task is completed.

You will start with the fundamentals of the Claude Agent SDK and progressively explore the capabilities required to build reliable AI agents.

You will learn how to:

• Set up and configure the Claude Agent SDK with TypeScript
• Understand the agent loop and autonomous task execution
• Work with built-in tools such as Read, Edit, Glob, Bash and WebSearchControl agent capabilities using allowed tools and permission modes
• Implement human approval for sensitive actions
• Create reusable configuration with SDK options and system prompts
• Build multi-turn agents using persistent sessions
• Stream agent reasoning, tool calls and results in real time
• Use hooks to intercept and control agent activity
• Create custom tools for your applications
• Connect external systems using MCP servers
• Integrate databases, APIs, browsers and development tools
• Use subagents and specialist agents for complex workflows

Read the full description on Udemy on Udemy (opens in a new tab)

Prices, discounts and availability are set by Udemy. We may earn a commission when you purchase through links on this site.