Feedback Pipeline MCP Server
Bridges Notion custom agents with automated pipelines to reproduce bugs and build fixes, enabling automated status updates, agent thread callbacks, and task page commenting.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Feedback Pipeline MCP Serverreproduce the login error and build a fix for task page 8c21"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Feedback Pipeline MCP Server
An MCP (Model Context Protocol) server that bridges Notion custom agents with Claude Code for automated bug reproduction and fix building.
Overview
This server allows a Notion custom agent to:
Reproduce bugs - Uses Playwright browser automation to reproduce reported bugs and capture screenshots
Build fixes - Uses Claude Code to implement fixes, run tests, and verify with Playwright
Related MCP server: Cloudflare Playwright MCP Example
Setup
Install dependencies:
npm installCreate a
.envfile with your Notion API token:NOTION_API_TOKEN=your_token_hereStart the server:
node server.jsExpose via ngrok (for Notion to connect):
ngrok http 3001
Architecture
Notion Custom Agent
│
▼ (via ngrok)
MCP Server (port 3001)
│
▼
Claude Code CLI
│
▼
Playwright MCP
│
▼
Notion Dev Server (port 3030)MCP Tools
reproduce_bug
Reproduce a bug using Playwright browser automation.
Parameters:
bug_description(required): Description of the bugreproduction_steps(optional): Steps to reproducecallback_agent_id(required): Agent ID for callbackcallback_thread_id(required): Thread ID for callbackcallback_page_id(required): Page ID for comment notification
build_fix
Build a fix or feature using Claude Code.
Parameters:
description(required): What to build or fixplan(optional): Implementation planverify_with_playwright(optional): Whether to verify with Playwright (default: true)callback_agent_id(required): Agent ID for callbackcallback_thread_id(required): Thread ID for callbackcallback_page_id(required): Page ID for comment notification
get_task_status
Check the status of a running or completed task.
Parameters:
task_id(required): The task ID to check
ping
Test connectivity - returns "pong".
Endpoints
GET /- SSE endpoint for MCP connectionPOST /- MCP message handlerGET /health- Health checkGET /tasks- List all tasksGET /tasks/:taskId- Get task detailsGET /demos- List demo recordingsGET /demos/:taskId/- Live progress page with screenshots
Callbacks
When a task completes, the server:
Posts results to the agent thread via Notion Agents SDK API
Adds a comment to the task page with demo URL and summary
Configuration
The server expects:
Notion dev server running on port 3030
Playwright MCP configured in the Claude Code environment
NOTION_API_TOKENenvironment variable for callbacks
This server cannot be deployed
Maintenance
Related MCP Connectors
Browser-backed QA with evidence and fix-ready reports for coding agents.
Browser-based QA for AI-built software. Test pages with real browsers via agents.
BugBug workflows for AI agents workflows for AI agents
- openhelmOAuthai.openhelm
Autonomous cloud agent tasks: real browser + your tools, structured evidence-backed results.
Related MCP Servers
- FlicenseNot gradedqualityCmaintenanceEnables AI assistants to control a browser through Playwright automation tools, allowing them to perform web tasks like navigation, typing, clicking, and taking screenshots. Deployed on Cloudflare Workers and compatible with various AI platforms including Claude Desktop and VS Code.-
- FlicenseNot gradedqualityCmaintenanceEnables AI assistants to control a web browser via Playwright on Cloudflare Workers for automated tasks like navigation, clicking, typing, and taking screenshots. It integrates with platforms like Claude Desktop and GitHub Copilot to perform web automation through a standardized toolset.-
- AlicenseNot gradedqualityBmaintenanceAn agentic QA framework that authors, generates, triages, and self-heals Playwright tests for any web app, usable from Claude Code/Desktop as an MCP server or from CI as a CLI.5MIT
- AlicenseNot gradedqualityBmaintenanceAutonomous QA platform powered by Claude + Playwright that allows AI to write, run, and fix tests for any project.MIT