Browser MCP
An MCP server that drives your already open Chrome and stops to ask you about a 2FA code, a CAPTCHA, or a choice only you can make
- 62 ★ on the repository
- Last commit
- MIT
Broad access
Why this level
- Drives your browser with accounts already signed in
- Can act on external sites on your behalf
Install
- Works in
Manual install
claude mcp add --scope user browser-mcp -- npx @agent360/browser-mcp@latestAfter the command, install the Chrome extension from the Chrome Web Store or load it unpacked, then restart the client.
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- Drives your browser with accounts already signed in
- Can act on external sites on your behalf
This is third-party code. Review the repository files before installing.
What it does
Browser MCP is a Chrome extension paired with a local MCP server. The agent acts in the same browser where you are already signed in, so it works on sites with login walls, two-factor prompts and bot protection. When it needs a code from your inbox, a CAPTCHA, or consent for a sensitive step, it hands control back to you and waits. Several agents can run at once, each in its own tab group. Everything runs on your machine, with no account and nothing sent to the authors.
Who it is for. For developers and QA engineers who need to automate work in a real, already signed-in browser.
Good fit when
- You need to act on a site where you are already signed in, with no API to wire up
- You need to pass a login with an email code or CAPTCHA while keeping the sensitive steps yourself
- You need an end-to-end run through your own web app as a real user
Not a fit when
- A stable service API exists and is more reliable than browser automation
- You cannot give the agent access to a browser holding your live sessions
Example request
Open my dashboard tab, pull this month's numbers into a table, I will confirm the login code from my inbox myselfMCP
- Transport
- stdio
- Authentication
- not required
Limitations
You need both halves: the Chrome extension and the local server, and one without the other does nothing. Chrome will not install the extension by itself, so you load it once by hand or from the Chrome Web Store. The agent acts as you inside your sessions, so sensitive actions are worth confirming yourself.
How to disable. Remove the browser-mcp server from the agent's MCP config and disable or remove the extension in chrome://extensions.
FAQ
How is this better than headless automation?
Actions run in your real Chrome with your sessions, so they pass where headless hits a login wall, CAPTCHA or bot protection.
Does data go to the authors?
No, the server runs locally, no account is needed, and the authors receive nothing.
Discussion
No questions or reviews yet.
Related
MCP serversAn MCP server and Chrome extension that let the agent drive your already open browser with your logged-in sessions, all locally
A Chrome extension with an MCP server: the agent drives your everyday browser with its sessions, tabs, bookmarks and history
A browser extension and local proxy that bring MCP tools into web chat UIs like ChatGPT, Perplexity and Grok
A local browser automation platform: MCP server, CLI and skills so an agent can drive Chrome, fill forms and pull data from pages
An MCP wrapper over the browser-use engine: a single tool drives a browser from a natural-language task.
An MCP bridge to control your browser from a coding agent: a Chrome extension and a set of tab and page automation tools
README in short
The README describes a two-part tool: a Chrome extension and a local MCP server on npm. The core idea is that the agent works in your real, already signed-in browser, so it handles sites where ordinary automation gets blocked. It highlights the ability to stop and ask the human for a 2FA code or CAPTCHA. It supports concurrent sessions with color-coded tab groups and several built-in service integrations. Install is a claude mcp add command plus loading the extension by hand.