Runbook AI

Provides browser automation via a Chrome extension by exposing an MCP server for live browser interactions.

javascript

0

GitHub Stars

javascript

Language

4 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "runbook-ai-runbook-ai-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "runbook-ai-mcp@latest"
      ]
    }
  }
}

You can run a browser-embedded MCP server that enables browser automation through a Chrome extension. This server lets terminal-based agents interact with live web pages by communicating with a local browser session, providing a practical way to automate tasks across websites while keeping your data in the browser.

How to use

You will operate the Runbook AI MCP Server from your development or agent environment. Start the server locally and connect an MCP client to it. The server talks to the Chrome extension in your browser via a WebSocket, then relays results back to your MCP client. Use this setup to guide AI agents to browse, extract information, fill forms, or navigate multi-page tasks on real websites without exposing your full browser history to third parties.

How to install

Prerequisites you need before installing this server are a working Node.js environment and a browser with the Runbook AI Chrome extension available in the Chrome Web Store.

Configure the MCP server in your client settings by adding the following JSON snippet to define the Runbook AI MCP server.

{
  "mcpServers": {
    "runbook-ai": {
      "command": "npx",
      "args": ["-y", "runbook-ai-mcp@latest"]
    }
  }
}

Additional configuration and usage notes

Install the Runbook AI browser extension from the Chrome Web Store and enable MCP in the extension settings. Configure your LLM API key and model name within the extension, using the recommended Gemini 3 Flash model when available. You can limit site access per your security preferences by adjusting the Site access settings in the browser extension details.

Start the MCP server from your client. The server will automatically start when invoked by your MCP client. Keep the extension side panel open while you run tasks.

Tool schema

The server provides a single tool to run tasks in the Chrome browser with AI and automation capabilities.

Tool name: browser-agent. This tool accepts a prompt describing the task you want the AI agent to execute.

Example task prompt: Go to google.com and search for 'MCP protocol'.

Available tools

browser-agent

Run a browser automation task using AI to control the Chrome extension and perform actions in the current browser session.