Skip to main content

​
Overview

The Anthropic Python SDK provides native support for MCP (Model Context Protocol) servers. This allows you to connect Claude directly to Observee’s tool ecosystem without additional frameworks.

​
Prerequisites

Before getting started, ensure you have:
  • Python 3.8 or later installed
  • Required packages installed: 
    pip install anthropic python-dotenv rich
  • Your Observee client ID for identifying your customer
  • Your Observee API key for authentication
  • Anthropic API key

​
Example

Connect Claude to Observee’s MCP server and access all available tools: 
import anthropic
from rich import print
from dotenv import load_dotenv
import os

load_dotenv()

# Your Observee MCP server URL
# Replace {your_client_id} with your actual client ID from the Observee dashboard
url = 'https://mcp.observee.ai/mcp?client_id={your_client_id}'

client = anthropic.Anthropic()

# Create streaming response with MCP server connection
stream = client.beta.messages.create(
    model="claude-sonnet-4-20250514",
    max_tokens=1000,
    messages=[{"role": "user", "content": "Check my latest email"}],
    mcp_servers=[
        {
            "type": "url",
            "url": url,
            "name": "observee",
            "authorization_token": "{observee_api_key}"  # Replace with your API key
        }
    ],
    extra_headers={
        "anthropic-beta": "mcp-client-2025-04-04"
    },
    stream=True
)

# Process the streaming response
print("Streaming response:")
for chunk in stream:
    if chunk.type == "content_block_delta":
        if chunk.delta.type == "text_delta":
            print(chunk.delta.text, end="", flush=True)
    elif chunk.type == "message_stop":
        print("\n\nStream completed.")
        break

print()  # Final newline for clean output

​
Learn More