Rod MCP Server

What is Rod MCP Server?

Rod MCP Server is a browser automation server built on the Rod library, providing tools for large language models to interact with web pages. It enables actions like clicking elements, taking screenshots, and saving pages as PDFs, running in headless or GUI mode.

How to use Rod MCP Server?

Clone the repository and build the binary (or download a release), then configure your MCP client with the command rod-mcp and arguments ["-c", "rod-mcp.yaml"]. The YAML file supports options for browser path, headless mode, temporary directory, sandbox, and proxy.

Key features of Rod MCP Server

• Browser automation via Go‑Rod
• Element clicking, screenshot capture, PDF generation
• Headless and GUI mode support
• High performance and stability
• Easy configuration and extension
• Designed for LLM interaction

Use cases of Rod MCP Server

• Automate web form filling under LLM guidance
• Capture page screenshots or PDFs for documentation
• Scrape or interact with dynamic web content
• Test web interfaces via language model commands

FAQ from Rod MCP Server

What are the prerequisites to run the server?

You need Go 1.23 or higher and a Chrome/Chromium browser installed on your system.

Can the browser run headlessly?

Yes. Set headless: true in the configuration file to run without a visible window.

How do I specify a custom browser path?

Use the browserBinPath option in the YAML config. If left empty, the system default browser is used.

Is sandbox mode configurable?

Yes. The noSandbox option lets you disable sandbox mode when needed.

What proxy types are supported?

The proxy setting supports SOCKS5 proxies for routing browser traffic.