2026 Q1 正式更名

Claude Agent SDK Developer Guide

The Claude Code SDK was officially renamed to Claude Agent SDK, offering a more complete AI Agent building experience. This guide covers Python/TypeScript quickstarts, core concepts, and enterprise best practices.

Python & TypeScript Renamed Q1 2026 QCode.cc Compatible

What is Claude Agent SDK

seo_agent_sdk.rename_title

Claude Code SDK Claude Agent SDK

seo_agent_sdk.rename_desc

seo_agent_sdk.why_rename_title

seo_agent_sdk.why_rename_desc

Unified Tool Interface

Shares the same tool system and Agent loop as Claude Code, enabling seamless switching between local and cloud deployments.

Production Ready

Built-in error retry, session management, streaming output and other enterprise features work out of the box.

Multi-language Support

Official Python and TypeScript SDKs with symmetric API design and comprehensive documentation.

Core Concepts

Tools

Functions callable by the Agent, declared with decorators that auto-generate JSON Schema for Claude to understand and invoke.

Agent Loop

Claude thinks → selects tool → executes → observes result → continues thinking, until the task is complete.

Memory / Context

Unified management of in-session short-term memory (conversation history) and cross-session long-term memory (file/database persistence).

Permissions

Fine-grained permission control: filesystem, network, code execution, and external API access scopes can all be configured independently.

Python Quickstart

Build your first Claude Agent in three steps, fully compatible with QCode.cc proxy.

Step 1: Install SDK

pip install anthropic-agent-sdk

Step 2: Create Agent

from anthropic_agent_sdk import Agent, tool

@tool
def get_weather(city: str) -> str:
    """Get current weather for a city"""
    return f"Weather in {city}: 22°C, sunny"

agent = Agent(
    model="claude-opus-4-7",
    tools=[get_weather],
    base_url="https://api.qcode.cc",  # QCode.cc proxy
    api_key="your-api-key",
)

result = agent.run("What's the weather in Beijing and Shanghai?")
print(result.output)

Step 3: Define Custom Tools

from anthropic_agent_sdk import Agent, tool
import requests

@tool
def search_web(query: str, max_results: int = 5) -> list[dict]:
    """Search the web and return results"""
    # Your search implementation here
    response = requests.get(f"https://api.search.example.com?q={query}&n={max_results}")
    return response.json()["results"]

@tool
def read_file(path: str) -> str:
    """Read contents of a local file"""
    with open(path, "r") as f:
        return f.read()

agent = Agent(
    model="claude-opus-4-7",
    tools=[search_web, read_file],
    base_url="https://api.qcode.cc",
    api_key="your-api-key",
    system="You are a helpful research assistant.",
)

result = agent.run("Search for recent news about Claude Agent SDK and summarize")
print(result.output)

TypeScript Quickstart

Claude Agent development in Node.js / Bun with full type support.

Step 1: Install SDK

npm install @anthropic-ai/agent-sdk

Step 2: Create Agent

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

const getWeather = tool({
  name: 'get_weather',
  description: 'Get current weather for a city',
  parameters: { city: { type: 'string' } },
  execute: async ({ city }) => `Weather in ${city}: 22°C, sunny`,
});

const agent = new Agent({
  model: 'claude-opus-4-7',
  tools: [getWeather],
  baseURL: 'https://api.qcode.cc',
  apiKey: process.env.ANTHROPIC_API_KEY,
});

const result = await agent.run("What's the weather in Beijing?");
console.log(result.output);

Step 3: Full Example

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

const searchWeb = tool({
  name: 'search_web',
  description: 'Search the web for information',
  parameters: {
    query: { type: 'string', description: 'Search query' },
    maxResults: { type: 'number', description: 'Max results', default: 5 },
  },
  execute: async ({ query, maxResults = 5 }) => {
    const res = await fetch(`https://api.search.example.com?q=${query}&n=${maxResults}`);
    const data = await res.json();
    return data.results;
  },
});

const agent = new Agent({
  model: 'claude-opus-4-7',
  tools: [searchWeb],
  baseURL: 'https://api.qcode.cc',
  apiKey: process.env.ANTHROPIC_API_KEY,
  systemPrompt: 'You are a helpful research assistant.',
  maxTurns: 10,
});

// Streaming support
for await (const event of agent.stream("Research the latest AI agent frameworks")) {
  if (event.type === 'text') process.stdout.write(event.text);
}

Agent SDK vs Managed Agents

seo_agent_sdk.vs_col_feature	seo_agent_sdk.vs_col_sdk	seo_agent_sdk.vs_col_managed
seo_agent_sdk.vs_row1_feature	Self-hosted (local / own server)	Anthropic cloud-hosted
seo_agent_sdk.vs_row2_feature	Manual instance management	Auto elastic scaling
seo_agent_sdk.vs_row3_feature	Prototyping, highly custom scenarios	Enterprise production Agent deployment

seo_agent_sdk.vs_managed_link

Using via QCode.cc

Set ANTHROPIC_BASE_URL to the QCode.cc endpoint and all Agent SDK features work reliably within China, no VPN required.

.env

# Python SDK
ANTHROPIC_BASE_URL=https://api.qcode.cc
ANTHROPIC_API_KEY=your-qcode-api-key

# TypeScript SDK
ANTHROPIC_BASE_URL=https://api.qcode.cc
ANTHROPIC_API_KEY=your-qcode-api-key

# Or set inline
agent = Agent(
    base_url="https://api.qcode.cc",
    api_key="your-qcode-api-key",
)

All Claude models (Opus 4.7 / Sonnet 4.6 / Haiku 4.5) available

Full tool calling (Function Calling) support

Streaming output support

Fully compatible with Managed Agents API

Start Building AI Agents Today

seo_agent_sdk.cta_btn_signup seo_agent_sdk.cta_btn_pricing