agent-browser-mcp
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 | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| get_setup_statusA | Return extension path, bridge ports, and connection status for setup/diagnostics. |
| list_tabsB | List currently connected browser tabs/sessions. |
| switch_tabB | Set the active browser tab by session id or URL substring. |
| open_urlB | Navigate the current tab to a URL using real-browser JS navigation. |
| open_new_tabA | Open a new browser tab with the given URL. |
| extension_pathA | Get absolute path to the unpacked Chrome extension directory for manual installation. |
| list_extensionsB | List Chrome extensions visible to the CDP bridge extension itself. |
| scan_pageA | Read the current page as simplified HTML/text, preserving login state from the real browser. |
| execute_jsC | Execute arbitrary JS in the current page context or send JSON CDP bridge commands through the page bridge. |
| cdp_commandC | Call a single Chrome DevTools Protocol command on the current or specified tab. |
| cdp_batchC | Run a CDP bridge batch command; pass the full JSON command object as text. |
| get_cookiesC | Get cookies for the current page or specified tab via the Chrome extension bridge. |
| capture_page_screenshotC | Capture a screenshot of the current page/tab via CDP and optionally save it to a file path. |
| capture_desktop_screenshotB | Take a desktop screenshot of the whole screen using mss; useful for physical-input verification. |
| mouse_moveB | Move the real mouse cursor to screen coordinates. |
| mouse_clickC | Click on the real desktop at screen coordinates. |
| mouse_dragC | Drag the real mouse from one point to another. |
| type_textB | Type text via the real keyboard, optionally after clicking a field. |
| hotkeyB | Send a hotkey chord like 'command,l' or 'ctrl,shift,p' via the real keyboard. |
| pointer_infoA | Report the current desktop mouse position and primary screen size. |
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 20 tools
Most tools have distinct purposes, but some overlap exists: 'capture_desktop_screenshot' and 'capture_page_screenshot' could be confused for similar screenshot tasks, and 'cdp_batch' and 'cdp_command' both handle CDP commands with subtle differences. The descriptions help clarify, but an agent might occasionally misselect between these pairs.
All tool names follow a consistent snake_case verb_noun pattern, such as 'capture_desktop_screenshot', 'list_tabs', and 'open_new_tab'. This predictability makes it easy for agents to understand and use the tools without confusion from mixed naming conventions.
With 20 tools, the count is slightly high but reasonable for a browser automation server that covers desktop interaction, CDP commands, and page management. It feels comprehensive without being overly bloated, though it borders on the upper limit of typical scoping.
The tool set provides complete coverage for browser automation, including navigation, tab management, input simulation, screenshot capture, CDP access, and diagnostics. There are no obvious gaps; agents can perform core workflows like opening URLs, interacting with pages, and debugging without dead ends.