Skip to main content

CrewAI

Connect a CrewAI agent to Civic using MCPClient with HTTP transport. CrewAI's MCP integration wraps discovered tools with MCPToolWrapper to make them usable by CrewAI agents.

Prerequisites​

  • Python 3.11+
  • A Civic account at app.civic.com with a configured toolkit
  • A Civic token and an Anthropic API key

Installation​

pip install crewai anthropic python-dotenv

Environment Variables​

CIVIC_URL=https://app.civic.com/hub/mcp?profile=your-toolkit
CIVIC_TOKEN=your-civic-token
ANTHROPIC_API_KEY=your-anthropic-key
Get Your Credentials

How to generate a Civic token and configure toolkit URL parameters

Connecting to Civic​

Use MCPClient with HTTPTransport to connect, then wrap the discovered tools with MCPToolWrapper:

import os
import asyncio
from dotenv import load_dotenv
from crewai import Agent, Task, Crew, LLM
from crewai.mcp.client import MCPClient, HTTPTransport
from crewai.tools.mcp_tool_wrapper import MCPToolWrapper

load_dotenv()

CIVIC_URL = os.environ["CIVIC_URL"]
CIVIC_TOKEN = os.environ["CIVIC_TOKEN"]

SERVER_PARAMS = {
    "url": CIVIC_URL,
    "headers": {"Authorization": f"Bearer {CIVIC_TOKEN}"},
    "transport": "streamable-http",
}

async def get_civic_tools():
    transport = HTTPTransport(
        url=CIVIC_URL,
        headers={"Authorization": f"Bearer {CIVIC_TOKEN}"},
        streamable=True,
    )
    async with MCPClient(transport=transport) as client:
        raw_tools = await client.list_tools()
    return [
        MCPToolWrapper(
            mcp_server_params=SERVER_PARAMS,
            tool_name=t["name"],
            tool_schema=t.get("inputSchema", {}),
            server_name="civic",
        )
        for t in raw_tools
    ]
note

MCPServerAdapter was removed in CrewAI v1.6.1. Use MCPClient + HTTPTransport + MCPToolWrapper as shown above.

Running the Agent​

async def main():
    tools = await get_civic_tools()

    llm = LLM(
        model="anthropic/claude-sonnet-4-6",
        api_key=os.environ["ANTHROPIC_API_KEY"],
    )

    assistant = Agent(
        role="Personal Assistant",
        goal="Help the user with their tasks using available tools",
        backstory="You are a capable assistant with access to the user's connected services through Civic.",
        tools=tools,
        llm=llm,
    )

    task = Task(
        description="What events do I have today?",
        expected_output="A list of today's calendar events.",
        agent=assistant,
    )

    crew = Crew(agents=[assistant], tasks=[task])
    result = crew.kickoff()
    print(result)

asyncio.run(main())

Production Configuration​

For production agents, lock to a specific toolkit using the profile URL parameter:

CIVIC_URL=https://app.civic.com/hub/mcp?profile=your-production-toolkit

Reference Implementation​

crewai-reference-implementation-civic

Complete implementation with FastAPI chat UI and deployment guide

Next Steps​

Agent Deployment

Production deployment guide: profile locking, URL params, authentication

Guardrails

Constrain what tools your agent can call

Audit Trail

Query what your agent did via Civic Chat

Get Credentials

Token generation and URL parameter reference