all tools / browser
As of the latest check, Glimind tracks 200 browser AI agent (MCP) tools, of which 0 are healthy right now. The table below ranks them live-first by real-time reliability.
| tool | status | what it does | |
|---|---|---|---|
| tinypilot_mouse_event tiny-pilot/tinypilot-mcp | unknown | Mouse move/click/scroll — relative (0.0–1.0) or pixel_x/pixel_y | alternatives |
| resolve_occlusion_and_scroll Alpha-Park/genpark-visual-viewport-element-occlusion-resolver-skill | unknown | Resolves visual viewport element occlusion and performs scrolling to bring elements into view | alternatives |
| check_connection SoulChildTc/ai-bridge | unknown | 检查与浏览器的连接状态 | alternatives |
| read_tab hunglp97/tabpilot-mcp | unknown | Reads the content of the current tab with progressive token slicing, optionally scoped to a CSS selector. | alternatives |
| query_dom hunglp97/tabpilot-mcp | unknown | Queries the DOM for a CSS selector and returns minimal information (e.g., button states, inputs, badges). | alternatives |
| fill_matrix hunglp97/tabpilot-mcp | unknown | Clicks multiple rows in a survey grid with delays to defeat React batching. | alternatives |
| puppeteer_navigate twolven/mcp-server-puppeteer-py | unknown | Navigate to any URL in the browser. | alternatives |
| puppeteer_fill twolven/mcp-server-puppeteer-py | unknown | Fill out input fields. | alternatives |
| puppeteer_screenshot twolven/mcp-server-puppeteer-py | unknown | Capture screenshots of the entire page or specific elements. | alternatives |
| puppeteer_click twolven/mcp-server-puppeteer-py | unknown | Click elements on the page. | alternatives |
| puppeteer_evaluate twolven/mcp-server-puppeteer-py | unknown | Execute JavaScript in the browser console. | alternatives |
| screenshot martingeidobler/android-mcp-server | unknown | Take screenshot for visual analysis | alternatives |
| get_ui_tree martingeidobler/android-mcp-server | unknown | Get UI element hierarchy with bounds, text, resource IDs, and state | alternatives |
| tap martingeidobler/android-mcp-server | unknown | Tap at screen coordinates | alternatives |
| tap_element martingeidobler/android-mcp-server | unknown | Tap element by resource-id, text, or content-desc | alternatives |
| tap_and_wait martingeidobler/android-mcp-server | unknown | Tap element, wait for UI to settle, return new UI tree — single round trip | alternatives |
| long_press martingeidobler/android-mcp-server | unknown | Long press at coordinates (context menus, drag handles) | alternatives |
| double_tap martingeidobler/android-mcp-server | unknown | Double tap at coordinates | alternatives |
| multi_tap martingeidobler/android-mcp-server | unknown | Tap the same coordinates N times with a fixed interval (spam tapping) | alternatives |
| tap_sequence martingeidobler/android-mcp-server | unknown | Multi-step action chain: taps, waits, text input, key presses, swipes in any order | alternatives |
| type_text martingeidobler/android-mcp-server | unknown | Type text into focused input | alternatives |
| press_key martingeidobler/android-mcp-server | unknown | Press key (back, home, enter, tab, delete, menu, etc.) | alternatives |
| swipe martingeidobler/android-mcp-server | unknown | Swipe gesture between coordinates | alternatives |
| scroll_to_element martingeidobler/android-mcp-server | unknown | Scroll until element is visible | alternatives |
| wait_for_element martingeidobler/android-mcp-server | unknown | Wait for element to appear (with timeout) | alternatives |
| screenshot snapshot-site/snapshot-site-mcp | unknown | Capture a screenshot of a web page | alternatives |
| chrome_navigate andrewlwn77/undetected-chrome-mcp | unknown | Navigate to URLs with anti-detection | alternatives |
| chrome_screenshot andrewlwn77/undetected-chrome-mcp | unknown | Capture page or element screenshots | alternatives |
| chrome_click andrewlwn77/undetected-chrome-mcp | unknown | Click page elements | alternatives |
| chrome_fill andrewlwn77/undetected-chrome-mcp | unknown | Fill input fields | alternatives |
| chrome_select andrewlwn77/undetected-chrome-mcp | unknown | Select dropdown options | alternatives |
| chrome_hover andrewlwn77/undetected-chrome-mcp | unknown | Hover over elements | alternatives |
| chrome_evaluate andrewlwn77/undetected-chrome-mcp | unknown | Execute JavaScript code | alternatives |
| start_network_monitoring andrewlwn77/undetected-chrome-mcp | unknown | Begin capturing network requests and responses | alternatives |
| stop_network_monitoring andrewlwn77/undetected-chrome-mcp | unknown | Stop network traffic capture | alternatives |
| get_network_requests andrewlwn77/undetected-chrome-mcp | unknown | Retrieve captured requests with filtering | alternatives |
| get_response_body andrewlwn77/undetected-chrome-mcp | unknown | Extract specific response content | alternatives |
| get_network_summary andrewlwn77/undetected-chrome-mcp | unknown | Get network activity statistics | alternatives |
| tree ellmos-ai/open-compute-mcp | unknown | Return UIA elements and a one-shot observation ID for their coordinates. | alternatives |
| invoke ellmos-ai/open-compute-mcp | unknown | Invokes a UI element by its UIA accessibility name (e.g., button press) | alternatives |
| get_screen_size ellmos-ai/open-compute-mcp | unknown | Virtual-desktop geometry + per-monitor breakdown (read-only). | alternatives |
| signal_show ellmos-ai/open-compute-mcp | unknown | Show the screen-usage signal overlay: glowing border + cursor ring colored per mode (control=red, observe=blue, …); persists in the server p | alternatives |
| signal_hide ellmos-ai/open-compute-mcp | unknown | Hide the signal overlay. | alternatives |
| signal_status ellmos-ai/open-compute-mcp | unknown | Overlay state + collect a pending abort-hotkey message (consumed on read). | alternatives |
| list_windows ellmos-ai/open-compute-mcp | unknown | List stable window/process IDs, exact titles. | alternatives |
| capture ellmos-ai/open-compute-mcp | unknown | Captures the current screen state and returns an observation ID for subsequent actions | alternatives |
| do ellmos-ai/open-compute-mcp | unknown | Executes a keyboard or mouse action on the desktop | alternatives |
| click_name ellmos-ai/open-compute-mcp | unknown | Clicks a UI element by its UIA accessibility name | alternatives |
| click shimondoodkin/screenmcp-claude-plugin | unknown | Clicks at specified coordinates in the screenshot space. | alternatives |
| double_click shimondoodkin/screenmcp-claude-plugin | unknown | Double-clicks at specified coordinates. | alternatives |
| drag shimondoodkin/screenmcp-claude-plugin | unknown | Drags from one point to another. | alternatives |
| scroll shimondoodkin/screenmcp-claude-plugin | unknown | Scrolls the mouse wheel. | alternatives |
| click arsonist-g/cdt | unknown | Click on an element in the page. | alternatives |
| navigate_page arsonist-g/cdt | unknown | Navigate the browser to a specified URL. | alternatives |
| take_snapshot arsonist-g/cdt | unknown | Take a snapshot of the current page. | alternatives |
| fill arsonist-g/cdt | unknown | Fill a form field with a value. | alternatives |
| describe_ui iksnerd/adb-mcp | unknown | Get the UI hierarchy with element centers in true device pixels | alternatives |
| tap iksnerd/adb-mcp | unknown | Tap at specified coordinates | alternatives |
| swipe iksnerd/adb-mcp | unknown | Swipe from one point to another | alternatives |
| run_sequence iksnerd/adb-mcp | unknown | Run a batch of steps with guards in one call | alternatives |
| verify_change iksnerd/adb-mcp | unknown | Check if a tap or key actually changed the UI | alternatives |
| tap_element iksnerd/adb-mcp | unknown | Tap on an element by ID | alternatives |
| long_press iksnerd/adb-mcp | unknown | Long press at specified coordinates | alternatives |
| key_combination iksnerd/adb-mcp | unknown | Send key combination (e.g., Ctrl+Alt+Del) | alternatives |
| drag iksnerd/adb-mcp | unknown | Drag from one point to another | alternatives |
| tap_on_text iksnerd/adb-mcp | unknown | Tap on an element with matching text | alternatives |
| pin_pad iksnerd/adb-mcp | unknown | Interact with a PIN pad | alternatives |
| extract_text The-Syntax-Slayer/browser-mcp | unknown | Extract text from any element | alternatives |
| fill_input The-Syntax-Slayer/browser-mcp | unknown | Type into input fields | alternatives |
| list_browsers The-Syntax-Slayer/browser-mcp | unknown | Detect installed browsers and profiles | alternatives |
| browse_url_cli The-Syntax-Slayer/browser-mcp | unknown | Headless CLI mode (no Playwright overhead) | alternatives |
| click_element The-Syntax-Slayer/browser-mcp | unknown | Click elements by CSS selector | alternatives |
| browser_bookmarks The-Syntax-Slayer/browser-mcp | unknown | Manage bookmarks across browsers (CRUD, search, sync, deduplication, tags, age analysis, link checking, export) | alternatives |
| browse_workflow The-Syntax-Slayer/browser-mcp | unknown | Multi-step agentic browsing from natural language task | alternatives |
| computer_screenshot ridelink0/claude-computer-use | unknown | Takes a screenshot of the screen or a specific region | alternatives |
| computer_apps ridelink0/claude-computer-use | unknown | Find the window, note its hwnd | alternatives |
| computer_click ridelink0/claude-computer-use | unknown | Clicks a UI element identified by its accessibility name or coordinates | alternatives |
| computer_type ridelink0/claude-computer-use | unknown | Types text into the currently focused UI element | alternatives |
| computer_snapshot ridelink0/claude-computer-use | unknown | Read a window's accessibility tree, every element gets an index | alternatives |
| browser_navigate zc-libre/camoufox-mcp-python | unknown | Navigate to a URL | alternatives |
| browser_navigate_back zc-libre/camoufox-mcp-python | unknown | Go back to the previous page | alternatives |
| browser_hover zc-libre/camoufox-mcp-python | unknown | Hover over an element | alternatives |
| browser_drag zc-libre/camoufox-mcp-python | unknown | Drag from one element to another | alternatives |
| browser_select_option zc-libre/camoufox-mcp-python | unknown | Select option(s) in a dropdown | alternatives |
| browser_type zc-libre/camoufox-mcp-python | unknown | Type text into an editable element (supports slow typing and submit) | alternatives |
| browser_fill_form zc-libre/camoufox-mcp-python | unknown | Fill multiple form fields in a single call | alternatives |
| browser_take_screenshot zc-libre/camoufox-mcp-python | unknown | Take a screenshot (viewport, full page, or element) | alternatives |
| browser_console_messages zc-libre/camoufox-mcp-python | unknown | Retrieve collected console messages | alternatives |
| browser_snapshot zc-libre/camoufox-mcp-python | unknown | Capture an accessibility snapshot of the current page | alternatives |
| browser_click zc-libre/camoufox-mcp-python | unknown | Click an element by ref (supports double-click, right-click, modifiers) | alternatives |
| browser_press_key zc-libre/camoufox-mcp-python | unknown | Press a keyboard key or key combination | alternatives |
| combo_render stranded7/agent-web-combo | unknown | Render a webpage using Kitesurf and optionally take a screenshot | alternatives |
| ax-tree @toolingtools/simview | unknown | Get the native accessibility tree (bypasses Metro) | alternatives |
| preview @toolingtools/simview | unknown | Open a live preview of a device | alternatives |
| tree @toolingtools/simview | unknown | Get the accessibility/inspection tree of the foreground app | alternatives |
| web_eval_agent Operative-Sh/web-eval-agent | unknown | Automated UX evaluator that drives the browser, captures screenshots, console & network logs, and returns a rich UX report. | alternatives |
| setup_browser_state Operative-Sh/web-eval-agent | unknown | Opens an interactive (non-headless) browser so you can sign in once; the saved cookies/local-storage are reused by subsequent web_eval_agent | alternatives |
| open_urls leynier/mcp-sys-bridge | unknown | Opens HTTP or HTTPS URLs in the default browser. URLs without a scheme use HTTPS. Other schemes are rejected. Preserves input order and dupl | alternatives |
| browser_get_cookies idugeni/a11ydriver-mcp | unknown | Get session cookies for active domain | alternatives |
| browser_list_tabs idugeni/a11ydriver-mcp | unknown | List all open Chrome tabs with IDs & titles | alternatives |
| browser_new_tab idugeni/a11ydriver-mcp | unknown | Open a new tab (optional url, groupId) | alternatives |
| browser_create_agent_group idugeni/a11ydriver-mcp | unknown | Create an isolated tab group for AI agent session | alternatives |
| browser_cleanup_agent_group idugeni/a11ydriver-mcp | unknown | Close all tabs inside an agent tab group after task finishes | alternatives |
| browser_switch_tab idugeni/a11ydriver-mcp | unknown | Switch active focused tab by tabId | alternatives |
| browser_close_tab idugeni/a11ydriver-mcp | unknown | Close specified tab by tabId | alternatives |
| browser_go_back idugeni/a11ydriver-mcp | unknown | Go back in browser history | alternatives |
| browser_go_forward idugeni/a11ydriver-mcp | unknown | Go forward in browser history | alternatives |
| browser_reload idugeni/a11ydriver-mcp | unknown | Reload current page | alternatives |
| browser_click idugeni/a11ydriver-mcp | unknown | Click element by coordinates (x, y) or element refId | alternatives |
| browser_hover idugeni/a11ydriver-mcp | unknown | Hover cursor over element or (x, y) coordinates | alternatives |
| browser_drag_and_drop idugeni/a11ydriver-mcp | unknown | Perform drag & drop from (fromX, fromY) to (toX, toY) | alternatives |
| browser_type idugeni/a11ydriver-mcp | unknown | Type text into element refId or active input | alternatives |
| browser_press_key idugeni/a11ydriver-mcp | unknown | Press special keys (Enter, Tab, Escape, Backspace, ArrowDown, ArrowUp) | alternatives |
| browser_scroll idugeni/a11ydriver-mcp | unknown | Scroll page vertically (deltaY) or horizontally (deltaX) | alternatives |
| browser_get_a11y_tree idugeni/a11ydriver-mcp | unknown | Get semantic Accessibility Tree snapshot with refId | alternatives |
| browser_evaluate_script idugeni/a11ydriver-mcp | unknown | Execute custom JS snippet on web context via CDP Runtime | alternatives |
| browser_screenshot idugeni/a11ydriver-mcp | unknown | Capture tab screenshot (supports fullPage: true) | alternatives |
| browser_print_to_pdf idugeni/a11ydriver-mcp | unknown | Print web page to PDF base64 document | alternatives |
| browser_navigate idugeni/a11ydriver-mcp | unknown | Navigate active tab to target url | alternatives |
| Click-Tool CursorTouch/Android-MCP | unknown | Click on the screen at the given coordinates. | alternatives |
| Long-Click-Tool CursorTouch/Android-MCP | unknown | Perform long click on the screen at the given coordinates. | alternatives |
| Type-Tool CursorTouch/Android-MCP | unknown | Type text on the specified coordinates (optionally clears existing text). | alternatives |
| Swipe-Tool CursorTouch/Android-MCP | unknown | Perform swipe from one point to another. | alternatives |
| browser_observe @hasna/browser | unknown | Build deterministic candidate actions for a task on a page | alternatives |
| browser_act @hasna/browser | unknown | Execute a deterministic action on a page | alternatives |
| browser_validate @hasna/browser | unknown | Validate a condition on a page | alternatives |
| browser_page_map @hasna/browser | unknown | Inspect a sanitized page model of a URL | alternatives |
| browser_video_start @hasna/browser | unknown | Start recording a browser session as video. | alternatives |
| browser_video_stop @hasna/browser | unknown | Stop an ongoing video recording. | alternatives |
| browser_video_delete @hasna/browser | unknown | Delete a recorded video. | alternatives |
| open_urls scitex-ai/scitex-scholar | unknown | Open URLs in browser and auto-import papers | alternatives |
| browser_snapshot huaka1/browser-mcp | unknown | Get page accessibility tree with ref IDs | alternatives |
| browser_scroll huaka1/browser-mcp | unknown | Scroll the page | alternatives |
| browser_back huaka1/browser-mcp | unknown | Go back in browser history | alternatives |
| browser_vision huaka1/browser-mcp | unknown | Take a screenshot of the page | alternatives |
| browser_get_images huaka1/browser-mcp | unknown | List images on the page | alternatives |
| browser_get_response_body huaka1/browser-mcp | unknown | Get full details of a request including headers and response body | alternatives |
| browser_intercept_request huaka1/browser-mcp | unknown | Intercept, mock, or block network requests | alternatives |
| browser_export_steps huaka1/browser-mcp | unknown | Export recorded steps as a jsonl test case and save login state | alternatives |
| browser_navigate huaka1/browser-mcp | unknown | Navigate to a URL and return page snapshot | alternatives |
| browser_click huaka1/browser-mcp | unknown | Click an element by ref ID | alternatives |
| browser_type huaka1/browser-mcp | unknown | Type text into an input field | alternatives |
| browser_press huaka1/browser-mcp | unknown | Press a keyboard key | alternatives |
| browser_get_network_logs huaka1/browser-mcp | unknown | Get network request logs with optional filters | alternatives |
| browser_list_steps huaka1/browser-mcp | unknown | List recorded operation steps in the current session buffer | alternatives |
| browser_console huaka1/browser-mcp | unknown | Get console output or execute JavaScript | alternatives |
| cnki_login ChengDSYJ-Studio/CNKI-Enhanced-MCP | unknown | 检测校园网/IP 权限，或打开浏览器等待账号、学校校外入口登录 | alternatives |
| app_window FrengkyS1/tauri-mcp | unknown | Resize / move / maximize / minimize / fullscreen the window; reports resulting size+position. | alternatives |
| app_execute_js FrengkyS1/tauri-mcp | unknown | Run arbitrary JavaScript in the window and return the result. | alternatives |
| app_type FrengkyS1/tauri-mcp | unknown | Type value into an input/textarea found by selector or text. | alternatives |
| app_hover FrengkyS1/tauri-mcp | unknown | Hover an element (by selector or text). | alternatives |
| app_select FrengkyS1/tauri-mcp | unknown | Pick an option from a native select. | alternatives |
| app_drag FrengkyS1/tauri-mcp | unknown | Drag an element onto another or by a pixel offset. | alternatives |
| app_read FrengkyS1/tauri-mcp | unknown | One element's live state: text, value, tag, displayed/enabled, optional attribute. | alternatives |
| app_wait_for FrengkyS1/tauri-mcp | unknown | Poll until a selector/text appears or disappears. | alternatives |
| app_console_logs FrengkyS1/tauri-mcp | unknown | Browser console output since launch. | alternatives |
| app_screenshot FrengkyS1/tauri-mcp | unknown | PNG screenshot of the window or a single element. | alternatives |
| app_list_windows FrengkyS1/tauri-mcp | unknown | List windows. | alternatives |
| app_switch_window FrengkyS1/tauri-mcp | unknown | Switch to another window. | alternatives |
| app_click FrengkyS1/tauri-mcp | unknown | Click by CSS selector or text. | alternatives |
| app_click_at FrengkyS1/tauri-mcp | unknown | Left/right/middle click at raw x,y viewport coordinates. | alternatives |
| app_scroll FrengkyS1/tauri-mcp | unknown | Mouse-wheel scroll by a pixel delta. | alternatives |
| app_press_key FrengkyS1/tauri-mcp | unknown | Raw keyboard: any key or modifier chord. | alternatives |
| live_resolve Patrizzos/StyleSpeak | unknown | Queries a running Chromium browser via CDP and returns exact computed styles and matched rules for a selector. | alternatives |
| get_active_tab_markdown Hasnat-GPU/OmniScrape MCP | unknown | Current tab as clean Markdown (Readability + Turndown, GFM tables). | alternatives |
| get_bridge_status Hasnat-GPU/OmniScrape MCP | unknown | Whether the browser is connected, and which tabs it can reach. | alternatives |
| cotest_act agentic-uitest | unknown | Perform an action (click, type, etc.) in the co-testing browser session. | alternatives |
| cotest_start agentic-uitest | unknown | Start a conversational co-testing session with a visible browser, allowing the agent to drive and the user to take over at any time. | alternatives |
| cotest_finalize agentic-uitest | unknown | Finalize the co-testing session, generating a Playwright spec and registering it as a test case. | alternatives |
| click_text TJCurnutte/mcp-computer-use | unknown | Click text found by OCR. | alternatives |
| browser_screenshot thekk1/playwright-mcp-guarded | unknown | Take a screenshot of the current page. | alternatives |
| browser_go_back thekk1/playwright-mcp-guarded | unknown | Go back to the previous page in the browser history. | alternatives |
| browser_get_html thekk1/playwright-mcp-guarded | unknown | Get HTML content of an element or the whole page. | alternatives |
| browser_get_attributes thekk1/playwright-mcp-guarded | unknown | Get attributes of an element. | alternatives |
| browser_get_console_messages thekk1/playwright-mcp-guarded | unknown | Get console messages from the page. | alternatives |
| browser_get_network_requests thekk1/playwright-mcp-guarded | unknown | Get network requests made by the page. | alternatives |
| browser_navigate thekk1/playwright-mcp-guarded | unknown | Navigate to a URL in the browser. | alternatives |
| browser_go_forward thekk1/playwright-mcp-guarded | unknown | Go forward to the next page in the browser history. | alternatives |
| browser_click thekk1/playwright-mcp-guarded | unknown | Click an element on the page. | alternatives |
| browser_fill thekk1/playwright-mcp-guarded | unknown | Fill a form field with text. | alternatives |
| browser_evaluate thekk1/playwright-mcp-guarded | unknown | Execute JavaScript in the page context. | alternatives |
| browser_get_text thekk1/playwright-mcp-guarded | unknown | Get text content of an element. | alternatives |
| browser_wait_for thekk1/playwright-mcp-guarded | unknown | Wait for a condition or element. | alternatives |
| browser_close thekk1/playwright-mcp-guarded | unknown | Close the current page or browser. | alternatives |
| runSEOAudit Browser Tools | unknown | Runs an SEO audit on the current page to evaluate search engine optimization. | alternatives |
| runBestPracticesAudit Browser Tools | unknown | Runs a best practices audit on the current page to check for general web development best practices. | alternatives |
| runPerformanceAudit Browser Tools | unknown | Runs a performance audit on the current page to identify bottlenecks and loading issues. | alternatives |
| runAccessibilityAudit Browser Tools | unknown | Runs an accessibility audit on the current page to check WCAG compliance and other accessibility issues. | alternatives |
| runNextJSAudit Browser Tools | unknown | Runs a NextJS-specific audit for best practices and SEO improvements. | alternatives |
| runDebuggerMode Browser Tools | unknown | Runs all debugging tools in sequence. | alternatives |
| runAuditMode Browser Tools | unknown | Runs all auditing tools in a sequence, including NextJS audit if framework is detected. | alternatives |
| pmux_browser_url yoway030/purplemux-mcp | unknown | Get or set the URL of a browser tab | alternatives |
| pmux_browser_screenshot yoway030/purplemux-mcp | unknown | Take a screenshot of a browser tab | alternatives |
| pmux_browser_console yoway030/purplemux-mcp | unknown | Get console logs from a browser tab | alternatives |
| pmux_browser_network yoway030/purplemux-mcp | unknown | Get network requests from a browser tab | alternatives |
| pmux_browser_network_body yoway030/purplemux-mcp | unknown | Get the body of a network request from a browser tab | alternatives |
| pmux_browser_eval yoway030/purplemux-mcp | unknown | Evaluate JavaScript in a browser tab | alternatives |
| flow_connect shaig-mahmudov/Google Flow Browser MCP | unknown | Attach to the running Chrome instance via CDP. | alternatives |
| flow_account_check shaig-mahmudov/Google Flow Browser MCP | unknown | Confirm the right Google account is signed in. | alternatives |
| browser_status mcp-patchright | unknown | Get browser status | alternatives |
As of the latest check, Glimind tracks 200 browser AI agent (MCP) tools, of which 0 are healthy right now. The table below ranks them live-first by real-time reliability. Glimind rates each neutrally (0–100) from safe liveness probes plus privacy-clean real-usage outcomes — see each tool's live page for its current score and a working alternative if it's down.
Ranked live-first by current verdict (healthy → degraded → down), then by the neutral reliability score. Glimind sells no tools, so the ranking is unconflicted.
Every tool here links to its live alternatives — capability-matched substitutes that are healthy now. Or query https://glimind.com/v1/alternatives/{toolId}.
Live data via MCP/REST. Neutral ratings — Glimind only measures.