Web Agent Qa

The MCP Server of operative.sh is an automated web debugging agent that can autonomously execute and debug web applications in the code editor, providing functions such as navigation, network request capture, and console error collection.

Developer tools Browser automation #Web debugging #Automated testing #Network monitoring #Error collection .Python

rating : 2 points

downloads : 5.1K

update time : 2025-04-28

Open Site

What is the web-eval-agent MCP Server?

The MCP Server launches an intelligent agent that autonomously tests and debugs web applications by simulating real user interactions. It integrates directly with your code editor to provide instant feedback.

How does it work?

After installation, the agent can be triggered to test web applications by navigating pages, clicking elements, and verifying functionality - all while capturing console logs and network traffic.

When should I use it?

Ideal for developers who want to automate testing of UI flows, verify API integrations, or debug complex frontend behavior without manual browser testing.

Key Features

Smart Navigation

Automatically browses your webapp using BrowserUse technology (2x faster with operative backend)

Network Traffic Capture

Intelligently filters and records all network requests for analysis

Error Monitoring

Captures console logs, warnings and errors during test execution

Autonomous Debugging

Cursor agent automatically tests if written code works as expected end-to-end

Advantages

Saves time by automating repetitive testing tasks

Provides comprehensive debugging data (network, console, screenshots)

Integrates directly with development workflow

Works with most modern web frameworks

Limitations

Initial setup requires some technical knowledge

May require adjustments for complex authentication flows

Currently better supported on macOS/Linux than Windows

Getting Started

Install prerequisites

Ensure you have brew, npm and jq installed

Run the installer

Download and execute the installation script (after getting a free API key)

Restart your IDE

Apply the changes by restarting your development environment

Start testing

Send prompts in chat mode to activate the web evaluation agent

Example Use Cases

Testing API Key Management

Verify the complete flow of creating and deleting API keys

Form Validation Check

Automatically test form submission with invalid data

Frequently Asked Questions

Why am I getting Playwright issues on first run?

How do I update the agent?

Is there a Windows version?

Additional Resources

Official Website

Get your API key and learn more about operative.sh

GitHub Repository

Source code and issue tracking

Star History

Project popularity over time

🚀 operative.sh Web Evaluation Agent MCP Server

Let the coding agent debug itself. You've got better things to do.

🚀 Quick Start

The MCP server of operative.sh enables direct execution and debugging of web applications in your code editor through a browser-based agent.

✨ Features

🌐 Navigate Your Web Application: Use BrowserUse (with the operative backend, 2x faster)
📊 Capture Network Traffic: Requests are intelligently filtered and returned to the context window
🚨 Collect Console Errors: Capture logs and error messages
🤖 Autonomous Debugging: The agent calls the web QA agent MCP server to test if the code works as expected

📦 Installation

macOS/Linux

Prerequisites: Install brew: /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)", npm: (brew install npm), and jq: brew install jq.
After obtaining your API key (free), run the installer:

# Feel free to review the installation script: curl -LSf https://operative.sh/install.sh | less -N
# Download, install, and remove the installation script
curl -LSf https://operative.sh/install.sh -o install.sh && bash install.sh && rm install.sh

Restart your favorite IDE to apply the changes.
Send prompts in chat mode to invoke the web evaluation tool!

Manual Windows Installation (Cline)

# 1. Get your API key on operative.sh
# 2. Install uv (curl -LsSf https://astral.sh/uv/install.sh | sh)
# 3. uvx --from git+https://github.com/Operative-Sh/web-eval-agent.git playwright install
# 4. Start the agent in Cline (you may need to restart Cline)

🔃 Update

Run uv clean.
Refresh the MCP server.

🚨 Important Note

An issue with the initial tool's invocation of Playwright has been fixed on April 14th.
If you encounter any problems, feel free to submit an issue in this repository!

📋 Change Log

April 29th - Agent overlay update - Pause/play/stop the agent's operation in the browser.

📋 Example of MCP Server Output Report

📊 Web evaluation report for http://localhost:5173 is complete!
📝 Task: Test the API key deletion process by navigating to the API key section, deleting a key, and evaluating the UX.

🔍 Agent Steps
  📍 1. Visit → http://localhost:5173
  📍 2. Click     “Login”        (Button index 0)
  📍 3. Fill in the username and password
  📍 4. Submit the form
  📍 5. Navigate to the API key management
  📍 6. Copy and delete the key
  📍 7. Verify the successful deletion

👁️  Check the “Operative Control Center” dashboard for real-time logs.