PatchworkMCP - Python

Drop-in feedback tool for Python MCP servers. Agents call this tool when they hit a limitation, and the feedback is sent to PatchworkMCP for review and action.

Setup

1. Go to patchworkmcp.com and create an account
2. Create a team and generate an API key
3. Configure your server (you'll need the server slug and API key)

Install

uv add httpx       # or: pip install httpx

Then copy feedback_tool.py into your project.

Configure

Set these environment variables (or pass them as arguments):

Usage

FastMCP

from mcp.server.fastmcp import FastMCP
from feedback_tool import register_feedback_tool

INSTRUCTIONS = """
If you encounter a limitation — a missing tool, incomplete data, wrong format,
or any gap that prevents you from fully completing the user's request — call
the `feedback` tool BEFORE responding to the user. Be specific about what you
needed and what would have helped.
"""

server = FastMCP("my-server", instructions=INSTRUCTIONS)
register_feedback_tool(server)

Raw MCP SDK

from feedback_tool import get_tool_definition, send_feedback

# In list_tools handler:
tools.append(get_tool_definition())

# In call_tool handler:
if name == "feedback":
    result = await send_feedback(arguments)

Django MCP (sync)

from feedback_tool import send_feedback_sync

result = send_feedback_sync(arguments)

Server Instructions

The instructions parameter on your MCP server is what tells agents to actually use the feedback tool. Without it, agents may see the tool but not know when to call it. The instruction text above is a good starting point — adapt it to your server's domain if needed.

How It Works

• Retries up to 2 times with exponential backoff (500ms, 1000ms)
• Retries on 429 (rate limit) and 5xx (server error) status codes
• On failure, logs the full payload with PATCHWORKMCP_UNSENT_FEEDBACK prefix for later replay
• Never throws or raises — always returns a user-facing message

License

MIT