mcpField
D
devtools-mcp

devtools-mcp

Local
github.com/stefanliBrowser Automation28Updated 73w ago
Repository

About devtools-mcp

devtools-mcp is an MCP server that gives LLMs access to the Chrome DevTools Protocol. It is intended for developers who want an AI assistant to drive a local Chrome instance, and it saves large binary responses such as screenshots and PDFs to files.

Key features

  1. 01Single cdp_command tool for any CDP command
  2. 02Binary responses such as screenshots saved to files
  3. 03File path references returned instead of raw data

Use cases

  1. 01Navigating a page and evaluating scripts through an AI assistant
  2. 02Capturing screenshots or PDFs to inspect page state

MCP support

This project is an MCP server that exposes a cdp_command tool, through which MCP clients send a CDP method name and parameters.

#devtools#chrome#llm-access

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

No install method listed yet

We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.

Repository

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

devtools-mcp

A MCP server that provides Chrome DevTools Protocol access to LLMs.

  • Execute any CDP command through the cdp_command tool
  • Automatic binary data handling - large responses (screenshots, PDFs) are saved to files

Setup

  1. Start Chrome with remote debugging:
/Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome \
  --remote-debugging-port=9222 \
  --remote-debugging-address=127.0.0.1 \
  --user-data-dir=/tmp/chrome-debug \
  --no-first-run \
  --disable-default-apps
  1. Install dependencies:
npm install
  1. Add as local MCP server. E.g. with Claude Code:
claude mcp add devtools-server -- npx tsx ~/projects/devtools-mcp/src/index.ts

Usage

The server exposes a single cdp_command tool that accepts:

  • method: Any CDP command (e.g., "Page.navigate", "Runtime.evaluate")
  • params: JSON string of command parameters (optional)

Binary data (screenshots, etc.) is automatically saved to ./cdp-output/ with file path references returned instead of raw data.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.