browser_click
Perform click on a web page
LABSCO SUMMARY
A page can be operated without anyone writing a selector.
browser_snapshot returns the page as an accessibility tree with a reference on each element, and every interaction call takes one of those references — so the agent aims at what the page exposes rather than at a CSS path that changes with the next deploy. When something goes wrong, the console logs and a screenshot are both one call away.
What it is
Twelve tools for a browser session: navigation and history, an accessibility snapshot that gives every element a reference, interaction aimed at those references, and two ways to see what happened — console logs and a screenshot.
What you get
- Navigation to a URL, and back and forward through history
- An accessibility snapshot of the page, which is where the element references every interaction call takes come from
- Clicks, hovers, typing with an optional submit, dropdown selection and key presses, each aimed at a named element and its reference
- A fixed wait in seconds for pages that need one
- The browser's console logs, and a screenshot of the current page
Requirements
npx on your PATH, and Chrome with the Browser MCP extension — the server drives your own browser through it, so a tab has to be connected (the extension's icon, then Connect) before any tool works. Interaction calls take an element and its reference, so browser_snapshot has to run before the first click.
Setup effort
Follow the published steps — Install the Browser MCP extension in Chrome (setup guide: https://docs.browsermcp.io)
USE CASES
01
Filling in a form on a site with no API
Snapshot the page for element references, then type and select against them. Nothing in the sequence depends on a CSS class that a redeploy can rename.
02
Getting the console error a user reported
Walk to the page the way they described, then read the console logs and screenshot what is on screen — the two things a bug report usually lacks.
03
Stepping through a signup or checkout flow
Move forward and back through the steps, waiting where the page needs a moment, and snapshot at each stage so the agent is acting on the current screen.
04
Checking that a page still works after a change
Load it, snapshot the tree, and compare the screenshot with what you expected. A broken render shows up in the snapshot before anyone reports it.
12 tools
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
browser_get_console_logs
Get the console logs from the browser
browser_go_back
Go back to the previous page
browser_go_forward
Go forward to the next page
browser_hover
Hover over element on page
browser_navigate
Navigate to a URL
browser_press_key
Press a key on the keyboard
browser_screenshot
Take a screenshot of the current page
browser_select_option
Select an option in a dropdown
browser_snapshot
Capture accessibility snapshot of the current page. Use this for getting references to elements to interact with.
browser_type
Type text into editable element
browser_wait
Wait for a specified time in seconds
BEFORE YOU INSTALL
What to know first
Tools you get
12 tools when we connected.
Requires
- npx on PATH
- Chrome with the Browser MCP extension installed, and the tab you want driven connected through it
Where it runs
The documented client runs the server on your machine.
CONTEXT COST
Light Standing cost
Rated by the size of the 12 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Connect your client
Server setup
Install the Browser MCP extension in Chrome (setup guide:
ENDPOINT URL
https://docs.browsermcp.ioAdd the config block to your MCP client
In Chrome, open the tab you want the agent to drive, click the Browser MCP extension icon in the toolbar and press Connect
MCP Clients
Any MCP client
ANY MCP CLIENT · JSON
{
"mcpServers": {
"browsermcp": {
"command": "npx",
"args": [
"-y",
"@browsermcp/mcp"
]
}
}
}
