Web Eval Agent

🚀 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 can directly execute and debug web applications autonomously in the code editor by launching a browser-based agent.

✨ Features

• 🌐 Navigate your web application Use BrowserUse (with a 2x speed boost from the operative backend).
• 📊 Capture network traffic Requests are intelligently filtered and returned to the context window.
• 🚨 Collect console errors Capture logs and errors.
• 🤖 Autonomous debugging The cursor agent calls the web QA agent MCP server to test if the code works as expected.

📦 Installation

macOS/Linux

1. Prerequisites:
   • Install brew: /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
   • Install npm: brew install npm
   • Install jq: brew install jq
2. After getting a free API Key, run the installer:

# View the installation script content as follows:
# curl -LSf https:// operative.sh/install.sh | less -N
# Download, install, and delete the installation script
curl -LSf https:// operative.sh/install.sh -o install.sh && bash install.sh && rm install.sh

3. Open your favorite IDE and restart it to apply the changes.
4. Send prompts in chat mode to invoke the web evaluation agent tool!

Manual Windows Installation (Command Line Interface)

# 1. Get your API Key at 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. Unleash the agent using the command line interface (you may need to restart Cline).

🔃 Update

• uv clean
• Refresh the MCP server.

⚠️ Important Note

• There were issues with the initial tool calling Playwright, and a fix was released on 4/14.
• If you have any problems, please open an issue in this repository.

📄 Change Log

• 4/29 - Agent overlay update - Pause/play/stop test runs in the browser.

💻 Usage Examples

Example MCP Server Output Report

# 🚀 operative.sh Web Evaluation Agent MCP Server

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

![Demo](./demo.gif)

## 🔥 Accelerate Your Debugging Experience

The MCP server of [operative.sh](https://www.operative.sh) can directly execute and debug web applications autonomously in the code editor by launching a browser-based agent.

## ⚡ Features

- 🌐 **Navigate your web application** Use BrowserUse (with a 2x speed boost from the operative backend).
- 📊 **Capture network traffic** Requests are intelligently filtered and returned to the context window.
- 🚨 **Collect console errors** Capture logs and errors.
- 🤖 **Autonomous debugging** The cursor agent calls the web QA agent MCP server to test if the code works as expected.

## 🏁 Quick Start (macOS/Linux)

1. Prerequisites: brew: `/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"`, npm: (`brew install npm`), jq: `brew install jq`
2. After getting a free API Key, run the installer:
```bash
# View the installation script content as follows:
# curl -LSf https:// operative.sh/install.sh | less -N
# Download, install, and delete the installation script
curl -LSf https:// operative.sh/install.sh -o install.sh && bash install.sh && rm install.sh

3. Open your favorite IDE and restart it to apply the changes.
4. Send prompts in chat mode to invoke the web evaluation agent tool!

🛠️ Manual Windows Installation (Command Line Interface)

# 1. Get your API Key at 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. Unleash the agent using the command line interface (you may need to restart Cline).

🔃 Update

• uv clean
• Refresh the MCP server.

🚨 Attention

• There were issues with the initial tool calling Playwright, and a fix was released on 4/14.
• If you have any problems, please open an issue in this repository.

Change Log

• 4/29 - Agent overlay update - Pause/play/stop test runs in the browser.

📋 Example MCP Server Output Report