Provides browser automation capabilities using Puppeteer to control a Chromium instance, allowing AI agents to navigate web pages, capture screenshots, evaluate JavaScript, and interact with DOM elements in development environments.
dev-chrome-monitor
A Model Context Protocol (MCP) server that keeps a Chromium instance attached with Puppeteer so tooling (Cursor, Trae, etc.) can inspect dev builds, capture screenshots, and drive the UI. The server exposes a permission-gated tool set so you remain in control of navigation, evaluation, and DOM automation.
Installation
Usage
Add the server to your project's mcp.json
:
Reload MCP in your IDE, then use the built-in tools:
dev_list_permissions
→ inspect which capabilities are enabled.dev_grant_permissions
/dev_revoke_permissions
→ grant or revoke permissions (acceptspermissions: ["status", "navigate", ...]
or["all"]
plusconfirm: true
).Other tools (
dev_status
,dev_screenshot
,dev_eval
, etc.) will only run after the matching permission is granted.
Environment Variables
MCP_DEV_URL
– default URL to open (the dev server you want to monitor).MCP_CHROME_PATH
– optional explicit Chrome/Chromium executable path.MCP_NAV_TIMEOUT
,MCP_WAIT_TIMEOUT
– navigation/wait timeouts in milliseconds.
Development
npm run build
copies the latest source (src/devChromeServer.mjs
) into
dist/
so the published package stays in sync.
License
MIT
This server cannot be installed
local-only server
The server can only run on the client's local machine because it depends on local resources.
Enables interaction with Chromium browser instances through Puppeteer for inspecting dev builds, capturing screenshots, and automating UI interactions. Features permission-gated tools for secure browser navigation, DOM manipulation, and JavaScript evaluation.