Claude Agent SDK Guide FAQs

Question 1

How do I manage agent tool permissions?

Accepted Answer

The skill demonstrates how to use explicit allowlists, blocklists, and strict permission modes to control which tools (such as Bash, Read, or Write) your agent is authorized to use.

Question 2

Can I use this with cloud providers like AWS or Google Cloud?

Accepted Answer

Yes, the guide covers authentication and configuration for Amazon Bedrock, Google Vertex AI, and Microsoft Foundry in addition to the native Anthropic API.

Question 3

What is the Claude Agent SDK?

Accepted Answer

The Claude Agent SDK is a development kit that allows you to embed the same autonomous agent infrastructure powering Claude Code into your own custom applications.

Question 4

Which programming languages are supported?

Accepted Answer

The SDK and this guide provide full support and comprehensive code examples for both TypeScript and Python environments.

Question 5

Does it support the Model Context Protocol (MCP)?

Accepted Answer

Yes, the guide includes implementation examples for extending your agent's capabilities using MCP servers to connect to external databases and custom tools.

Claude Agent SDK Guide

关于

主要功能

使用场景

Claude Agent SDK Guide

关于

主要功能

使用场景