How to Install Anthropic Claude MCP in OpenAI Agents SDK
Nest Claude within Claude — use Claude as a sub-agent for specialised reasoning tasks
What Anthropic Claude MCP does
The Anthropic Claude MCP server allows you to call Claude models as a tool from within another Claude session — enabling powerful multi-agent architectures where specialised sub-agents handle specific tasks. Features: - Call Claude Haiku, Sonnet, or Opus from within any MCP client - Pass custom system prompts to specialise each sub-agent - Stream responses for long-form generation tasks - Control temperature, max tokens, and other parameters - Chain multiple Claude calls in a single workflow - Use different models for different task types (fast Haiku for triage, Opus for reasoning) - Built-in prompt caching for cost efficiency Architecture patterns this enables: - Orchestrator + worker: main Claude delegates to specialised sub-agents - Critique and refine: one Claude generates, another reviews and improves - Parallel processing: multiple Claudes working on parts of a problem simultaneously - Specialised personas: legal Claude, code Claude, writing Claude all in one workflow
Installation steps
- 1Install: pip install openai-agents
- 2Copy the snippet below into your agent code
- 3Replace placeholder env values with your real keys
- 4Pass mcp_servers=[mcp_server] to your Agent()
Python snippet
from agents import Agent
from agents.mcp import MCPServerStdio
mcp_server = MCPServerStdio(
command="npx",
args=["-y", "@anthropic-ai/mcp"],
env={
"ANTHROPIC_API_KEY": "sk-ant-...",
},
)
agent = Agent(
name="My Agent",
model="gpt-4o",
mcp_servers=[mcp_server],
)Install in other clients
For maintainers
If you maintain Anthropic Claude MCP, add this badge to your README to show it's verified on CuratedMCP:
[](https://curatedmcp.com/marketplace/anthropic-claude-mcp)