Connect
CrewAI
CrewAI's MCPServerAdapter wraps any MCP server so every CrewAI agent can call it as a regular tool. Plug Dock in once and your crews can co-edit shared workspaces across sub-tasks.
Client
CrewAI
Python multi-agent orchestrator. MCP via crewai-tools MCPServerAdapter.
HTTP JSON-RPC (streamable-http)
Dock MCP
trydock.ai/api/mcp
37 tools · OAuth 2.1 + DCR · Bearer
Auth path
1
Mint a
dk_ key in Dock Settings → API keys.2
Paste it as
Authorization: Bearer dk_… in the client config.3
Client calls Dock MCP directly on every request.
Prerequisites
- Python 3.10+.
crewai-toolswith MCP support installed (pip install "crewai-tools[mcp]").- A Dock
dk_key.
crew.pypython
import os
from crewai import Agent, Crew, Task
from crewai_tools import MCPServerAdapter
dock_server = MCPServerAdapter({
"url": "https://trydock.ai/api/mcp",
"transport": "streamable-http",
"headers": {
"Authorization": f"Bearer {os.environ['DOCK_API_KEY']}"
},
})
with dock_server as dock_tools:
editor = Agent(
role="Dock editor",
goal="Keep content-pipeline tidy and moving.",
tools=dock_tools,
)
seal = Task(
description="Seal every drafted row older than 14 days.",
agent=editor,
)
Crew(agents=[editor], tasks=[seal]).kickoff()Verify
DOCK_API_KEY=dk_live_... python crew.py
# → Crew kicked off. Editor sealed 3 rows.Troubleshooting
Symptom
Fix
ImportError: MCPServerAdapter not found
You installed
crewai-tools without the MCP extra. Reinstall with pip install "crewai-tools[mcp]".Adapter context exits without listing tools
Some environments strip the trailing slash from URLs. Use the exact string
https://trydock.ai/api/mcp (no trailing slash, no path mangling).