Browser Control MCP by Nova (CIVAI)
Server Details
Control your Chrome from chat: open websites, read what's on the page, click buttons, and fill in f…
- Status
- Healthy
- Uptime
- 100.0% over the last 24 hours
- Last Tested
- Transport
- Streamable HTTP · MCP 2025-11-25
- URL
- ChatGPT
- –
- Claude
- –
TDQS
Scored across 26 tools
Several tools overlap heavily: browser_click, browser_interact, and browser_mouse_click_xy all click; browser_fill, browser_type, and browser_dom_input all enter text; and browser_actions is a catch-all batch that duplicates almost every individual action. The descriptions add hints (ref vs x/y, batch vs single) but an agent must reason carefully to pick correctly, and the presence of a generic batch tool blurs boundaries further.
Nearly everything is snake_case with a clear browser_ prefix for page actions and an unprefixed group for session/infrastructure tools, which reads as a deliberate two-tier convention. Minor deviations (browser_mouse_click_xy ordering, noun-style browser_actions/browser_tabs) are small and still readable.
At 26 tools this is on the heavy side for a browser controller, and a good chunk of the surface is redundant (multiple click/fill variants plus a batch tool). A tighter set of ~12-15 primitives plus the batch action would cover the same ground with less selection risk.
Coverage is broad: navigation, snapshots, clicking, typing, keys, scrolling, select, drag, JS execution, cookies, tabs, screenshots, plus session and license/identity management. Gaps exist (no explicit wait-for-selector/navigation, no hover, no file upload or tab-close primitive), but these are workable via browser_exec_js or browser_actions.
Available Tools
26 toolsbrowser_actionsAInspect
Low-level action batch (JSON array string). Pass tab_id when batch includes switch_tab, close_tab, mouse_click, scroll, or keys.
Inputs: actions_json (required), tab_id, timeout_ms, notification_message.
browser_clickCInspect
Click by snapshot ref (Cursor parity). Snapshot first.
Inputs: ref (required), tab_id (required), element, notification_message.
browser_dom_inputCInspect
Run dom input method chain (e.g. 2,7,0); use after list_input_methods on stubborn fields.
Inputs: text (required), tab_id (required), method, selector, +2 more.
browser_dragCInspect
Drag between viewport coordinates.
Inputs: endX (required), endY (required), startX (required), startY (required), +2 more.
browser_exec_jsAInspect
Run JavaScript on a target tab (same CDP evaluate path as BrowserGPT Quick Task exec_js). Requires tab_id from browser_tabs. Returns ok, value, logs, metrics.
Inputs: tab_id (required), javascript (required), timeout_ms, notification_message.
browser_fillCInspect
Fill field by ref.
Inputs: ref (required), value (required), tab_id (required), input_method, +1 more.
browser_get_bounding_boxCInspect
Get ref bounding box from last snapshot.
Inputs: ref (required), tab_id (required), notification_message.
browser_get_cookiesAInspect
Read session data for a tab via BrowserGPT extension: chrome.cookies (HttpOnly), optional localStorage/sessionStorage keys. Operator must approve the in-tab permission prompt. Use for session export / porting auth; do not paste values into chat.
Inputs: tab_id (required), cookie_domains (required), confirmation_message (required), cookie_names, +3 more.
browser_highlightCInspect
Briefly outline ref on page.
Inputs: ref (required), tab_id (required), notification_message.
browser_interactDInspect
Interact via snapshot ref.
Inputs: op (required), ref (required), tab_id (required), key, +4 more.
browser_mouse_click_xyBInspect
Click at viewport x/y when snapshot refs are empty.
Inputs: x (required), y (required), tab_id (required), notification_message.
browser_press_keyCInspect
Press keyboard key in tab.
Inputs: key (required), tab_id (required), code, modifiers_json, +1 more.
browser_screenshotBInspect
JPEG screenshot of the session viewport.
Inputs: tab_id (required), notification_message.
browser_scrollCInspect
Scroll wheel; optional ref scrolls element into view first.
Inputs: tab_id (required), ref, amount, deltaX, +2 more.
browser_select_optionCInspect
Select options on a select ref.
Inputs: ref (required), tab_id (required), values_json (required), notification_message.
browser_snapshotCInspect
Accessibility snapshot with refs.
Inputs: tab_id (required), interactive_only, notification_message.
browser_tabsBInspect
List tabs in the open control session.
Inputs: notification_message.
browser_typeCInspect
Type into field by ref.
Inputs: ref (required), text (required), tab_id (required), slowly, +2 more.
ensure_remote_runnerAInspect
Start/prepare Nova remote BrowserGPT runner (ensure-remote-browser). Call when list_browsers shows no online+ready remote node; then list_browsers again.
Inputs: notification_message.
get_network_identityAInspect
Where this user's requests to the site leave from (a VPN country, the server, or the app's default) and which browser they present. Read-only.
Inputs: no named parameters.
list_browsersBInspect
List BrowserGPT browser nodes (local + remote) for the configured license.
Inputs: notification_message.
list_input_methodsBInspect
List BrowserGPT dom input method ids; optional page_url for domain-specific hint.
Inputs: page_url, notification_message.
open_control_sessionCInspect
Open MCP control session on target browser node_id.
Inputs: target (required), notification_message.
set_browsergpt_licenseAInspect
Save BrowserGPT license when MCP headers did not already provide it (Authorization Bearer license). Optional when Nova/SecOps derived license from auth. Call before list_browsers only if license_not_set.
Inputs: license_key (required), notification_message.
set_network_identityAInspect
Choose where this user's requests to the site leave from and which browser they present, only when the user asks (e.g. 'connect through Nigeria', 'use my own user agent'). Sites may ask the user to sign in again when this changes. reset=true goes back to how they signed in.
Inputs: reset, route, locale, country, +4 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
Control a real browser from AI agents: navigate, click, type, screenshot, read the page.
Let AI agents use your real, signed-in Chrome: tabs, clicks, forms, screenshots and video.
AI-powered browser automation — navigate, click, fill forms, and extract data from any website.
Give AI agents a real cloud Chrome: open sessions, navigate, read pages, click, type and screenshot.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables AI assistants to control a Chrome browser through natural language, supporting navigation, interaction, tab management, and data extraction.MIT
- AlicenseNot gradedqualityCmaintenanceLets AI assistants control your real Chrome browser to perform web tasks like reading pages, taking screenshots, clicking, and typing, using your existing logged-in sessions.145MIT
- AlicenseNot gradedqualityBmaintenanceEnables AI assistants and terminal users to control a logged-in Chrome browser, performing actions like opening pages, searching, filling forms, and taking screenshots without re-authentication.MIT
- AlicenseNot gradedqualityDmaintenanceTurn your Chrome browser into your intelligent assistant - Let AI take control of your browser, transforming it into a powerful AI-controlled automation tool.MIT
Glama MCP Gateway
Add one secure layer between your agents and this server.