browser_close
Close the browser instance
LABSCO SUMMARY
A visible Chrome that reads the page before it clicks.
The content-first workflow is the design choice worth knowing: the server pushes you to read the page's HTML or text before interacting, which holds up better than deciding from a screenshot. Two tools do not do what their names suggest — solve_captcha is a placeholder that calls no solving service beyond a best-effort click on a detected Turnstile widget at startup, and save_content_as_markdown is advertised but blocked by the workflow validator, which the project records as a known bug. The anti-detection settings help; sites can still spot and block it.
What it is
A local Chrome driven through rebrowser-puppeteer-core with a small set of anti-detection settings. One browser session at a time, visible by default, with the exact Chrome process it started tracked so closing it never touches your own browser.
What you get
- A Chrome instance started with anti-detection settings and automatic Chrome path detection on Windows, macOS and Linux — headed by default, headless on request
- Proxy, a custom Chrome path, and a dedicated automation profile as launch options
- Page HTML or text read back — the documented first step before interacting, and steadier than a screenshot for deciding what to do next
- A CSS selector found from text you can actually see on the page, so you do not have to guess it
- Click, type with a per-character delay, wait on a condition, and random scrolling with natural timing
- Only its own Chrome process closed on shutdown
Requirements
Node.js 18 or newer, npm and npx, and Google Chrome or Chromium installed. Launched over stdio with npx -y puppeteer-real-browser-mcp-server@latest. Nothing to authorize — it drives a browser on your machine with your own permissions, which is also why logins, forms, purchases and downloads are worth reading before you approve the call. One browser session at a time.
Setup effort
One command — npx -y puppeteer-real-browser-mcp-server@latest
USE CASES
01
Reading a page a plain HTTP fetch cannot get
Start the browser, navigate, and take the HTML or text back rather than a screenshot.
02
Clicking something you can only describe
Find the selector from the text visible on the page, then click it and wait for the navigation.
03
Filling in a form
Type into each field with a delay, then wait on the condition that says it worked.
04
Reaching content that only loads as you scroll
Scroll with natural timing, then read the content again.
05
Keeping a session between runs
Start the browser against a dedicated automation profile so the login survives the next launch.
11 tools
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
browser_init
Initialize a new browser instance with anti-detection features and automatic Chrome path detection
click
Click on an element
find_selector
Find CSS selector for element containing specific text
get_content
Recommended method to get page content (HTML or text) - More reliable than screenshots for content analysis and navigation tasks
navigate
Navigate to a URL
random_scroll
Perform random scrolling with natural timing
save_content_as_markdown
Extract page content and save it as a formatted markdown file
solve_captcha
Attempt to solve CAPTCHAs (if supported)
type
Type text into an input field
wait
Wait for various conditions
BEFORE YOU INSTALL
What to know first
Tools you get
11 tools when we connected. solve_captcha is a placeholder that does not solve CAPTCHAs, and save_content_as_markdown is blocked by a known code issue, so 9 of the 11 work.
Requires
- Node.js 18 or newer
- npm and npx (included with the normal Node.js installer)
- Google Chrome or Chromium
- An MCP client that can start a local stdio server
CONTEXT COST
Light Standing cost
Rated by the size of the 11 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Connect your client
Server setup
Any MCP client
TERMINAL
npx -y puppeteer-real-browser-mcp-server@latestMCP Clients
Claude Desktop, Cursor
JSON
{
"mcpServers": {
"puppeteer-real-browser": {
"command": "npx",
"args": [
"-y",
"puppeteer-real-browser-mcp-server@latest"
]
}
}
}Claude Code
CLAUDE CODE · TERMINAL
claude mcp add --transport stdio puppeteer-real-browser \
-- npx -y puppeteer-real-browser-mcp-server@latest
