Browser MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| navigateC | Navigate to a URL in the browser |
| snapshotC | Get page text content |
| clickC | Click an element by CSS selector |
| screenshotC | Take screenshot (returns base64 PNG) |
| evaluateC | Execute JavaScript in page MAIN world |
| tabsB | List open browser tabs |
| connectB | Check if browser extension is connected |
| open_tabC | Open a new browser tab with a URL |
| close_tabC | Close a browser tab by tabId |
| switch_tabC | Switch to a specific tab (bring it to front) |
| typeC | Type text into the focused element via CDP Input.insertText |
| fillB | Fill an input/textarea by CSS selector, dispatching input/change events |
| press_keyC | Press a keyboard key (e.g. Enter, Escape, Tab, ArrowDown, Backspace) |
| wait_forB | Poll for a CSS selector to appear on the page (100ms interval, default 10s timeout) |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 14 tools
Most tools target clearly distinct actions (navigate, click, screenshot, wait_for, tab management). The main overlap is type vs fill, both of which insert text, but the descriptions differentiate them as focused-element CDP input vs selector-based event-dispatching fill. tabs vs open_tab/switch_tab/close_tab is also close but readably separated by descriptions.
All names are lowercase and mostly snake_case, but the convention is mixed: bare single-word verbs/nouns (navigate, snapshot, click, evaluate, tabs, connect) coexist with verb_object compounds (open_tab, close_tab, switch_tab, press_key, wait_for). It is readable but not a predictable verb_noun pattern throughout.
14 tools sit comfortably within a well-scoped range, and each maps to a real browser automation step (navigation, interaction, waiting, tabs, connectivity). A couple could arguably be folded together (type/fill, tabs family), but nothing feels padded.
Core automation lifecycle is covered: navigate, inspect (snapshot/screenshot/evaluate), interact (click/type/fill/press_key), wait, and tab management. Gaps remain for common flows such as scroll, back/forward/reload navigation, hover, and select-option handling, which agents may have to work around with evaluate.