Browser MCP vs Browserctl MCP
A side-by-side look at the Browser MCP and Browserctl MCP MCP servers: transport, auth, tools, quality signals, and the config to install either one in Claude, Cursor, Windsurf or VS Code.
Verdict: choose Browser MCP for browser automation work as a local process; choose Browserctl MCP for browser automation as a local process. You can also run both side by side in one mcpServers block.
Which MCP server should you choose?
Choose Browser MCP when
- You need dedicated capabilities in the Browser Automation domain.
- You prefer a local stdio subprocess transport.
- Its auth model fits you: No auth required.
- You want tools like
browser_navigate, browser_get_page_content, browser_extract_list.
Choose Browserctl MCP when
- You need dedicated capabilities in the Browser Automation domain.
- You prefer a local stdio subprocess transport.
- Its auth model fits you: No auth required.
- You want tools like
browser_load_tools, browser_list_available_tools, browser_action.
Feature & specification comparison
Rows where the two servers differ are marked.
| Spec | ||
|---|---|---|
| Category | Browser Automation | Browser Automation |
| Quality signal | 69/100 · Great | 63/100 · Good |
| Install command | Confirmed | Confirmed |
| Transport | Local subprocess (stdio) | Local subprocess (stdio) |
| Auth | No auth required | No auth required |
| Pricing | Free | Free |
| Required env vars | None listed | None listed |
| Verified owner | Community listing | Community listing |
| GitHub stars | 58 | 0 |
| Directory activity | 7 views 0 installs 0 upvotes | 1 views 0 installs 0 upvotes |
| Tools | 40 | 25 |
| Full listing | View Browser MCP | View Browserctl MCP |
Tools & capabilities
Browser MCP tools 40
browser_navigateNavigate the active browser tab to a URL. Reuses the current tab by default (no tab spam). Pass new_tab=true only when you need to keep the current page open. In the current tab the answer is ok:true only when Chrome took the address and the new page is proven loaded within about 25 s (the wait ends half a second before that limit). If the page asks "Leave site?", the default answer is Stay (an armed browser_handle_dialog answers instead) and the reply is error beforeunload-blocked with the tab's actual url: the page may hold unsaved work, so ask the user (browser_ask_user) before you call again with leave_page:true. Other errors, each with sent (was the address sent to Chrome?) and a note on what to do: navigation-in-progress (another navigate is working in this tab, or - with op_id null - an earlier answer to a dialog on the tab is not confirmed yet; wait a moment), dialog-open (a dialog is on screen - answer it with browser_handle_dialog), tab-not-responding (busy page, or a dialog Chrome does not show us), navigation-failed, dialog-unconfirmed, navigation-unconfirmed (no load proven in time; may_complete_later and pending_url say it can still happen - read the tab before sending it again; a download or a 204 answer never loads a page and ends here too), tab-closed (Chrome no longer has the tab), session-changed (another call changed the session's active tab while this one waited; nothing was sent).browser_get_page_contentGet the content of the current page as text or HTML. Pass `selector` to read one part of a large page instead of all of it - on a big logged-in app the full HTML can run past a million characters. The answer is capped at 30000 characters by default and says so when it had to cut, with the real length, so a truncated page is never mistaken for a whole one.browser_extract_listRead the rows of a long or virtualised list by scrolling its container until no new rows appear. Use this instead of browser_get_page_content whenever a page shows a repeating list longer than the viewport - mail lists (Outlook, Gmail), invoice/billing tables, search results, transaction histories. Those UIs keep only ~7 rows in the DOM at a time, so a single page read returns a sliver and looks complete. Pass the CSS selector of one repeating row (e.g. '[role="option"]', 'tr', '[role="listitem"]'); the scrollable ancestor is found automatically. When the tab is the active one in its window, it scrolls with real wheel events, also at the bottom, so a feed that loads more rows on wheel (Threads, X) loads them. A hidden page (a background tab) gets no wheel or scroll events, IntersectionObserver callbacks or animation frames, so a list that loads or renders rows while you scroll stops at what is already there; then reached_end is false and `note` says why. Returns deduplicated row text, scroll_method (wheel or script) and reached_end, which is true only when the list stood at its bottom in a visible page with no new rows, growth or movement for stable_rounds rounds in a row. If the selector matches no rows with text, the answer is error "no-rows".browser_screenshotTake a screenshot of the visible area of the current tab. Returns base64 PNG, or saves to disk if path is provided. The image is in device pixels; when Chrome reports them in time, the answer also gives the CSS viewport size and devicePixelRatio, which browser_click_xy coordinates need. Chrome stays where it is: if its window is covered by other windows, there is no picture and the answer is error "window-covered".browser_execute_scriptExecute JavaScript in the current page. IMPORTANT: the parameter is `code` (NOT `script` - though that alias is accepted), and it must be an EXPRESSION, not statements: use an IIFE `(() => { ...; return x; })()`. Top-level `return` is a syntax error (the handler wraps code in parentheses).browser_double_clickTrue double-click on an element (two trusted press/release pairs with escalating clickCount). Use for open-item actions (calendar events, file lists) where two single clicks would trigger inline-rename instead (e.g. OWA month view). If the mouse would not reach the target at its center - another element lies in front of it (an overlay, a dialog's backdrop), the target does not take clicks there (pointer-events, visibility, clipping), or a different control inside it would get the click (a delete button on a card; for a container that is not clickable itself, its own plain link with a real address (a hash route with a path, such as #/orders, counts; a fragment with no target on the page does not) does not count, nor does a web component's own button that covers it, nor a box with a checkbox, switch or radio role right next to a hidden field and the only one in the nearest label around it, a label of that field (a box inside a closed shadow root cannot be seen and is not counted), which counts as that field) - the answer is ok:false with error "covered" and covered_by. Found before the mouse is sent, nothing is sent; if it comes in front afterwards, the rest of the presses and clicks are stopped (a release the page moves elsewhere is let through), and the answer also has maybe_landed: true, because the page may already have reacted.browser_right_clickRight-click an element (trusted CDP mouse events) to open page-level context menus (web apps like OWA/Google Docs render their own). Note: Chrome's NATIVE context menu does not open via CDP - only in-page menus. If the mouse would not reach the target at its center - another element lies in front of it (an overlay, a dialog's backdrop), or the target does not take clicks there (pointer-events, visibility, clipping) - the answer is ok:false with error "covered" and covered_by. Found before the mouse is sent, nothing is sent; if it comes in front afterwards, the rest of the presses and clicks are stopped (a release the page moves elsewhere is let through), and the answer also has maybe_landed: true, because the page may already have reacted.browser_click_xyESCAPE HATCH: Click at raw viewport coordinates (CSS pixels) with fully trusted mouse events. Use when a visible button resists every selector strategy (Azure portal dialogs, Knockout-bound divs, canvas UIs): take a screenshot, read the button's position, click its center. Coordinates are CSS pixels, NOT image pixels: convert with the viewport line browser_screenshot returns: x_css = x / W * css_width and y_css = y / W * css_width, where (x, y) is the point in the image and W is the width the image is shown to you at.browser_reattach_debuggerRECOVERY: Force-detach and re-attach the Chrome debugger on the current tab. Use when interactive tools (click/fill/press_key) start timing out or reporting ghost-attach ("Debugger attach failed ... ghost") while list_tabs still works - faster than reloading the extension.browser_clickClick an element on the page. Supports CSS selectors AND text-based selectors. Auto-scrolls element into view. Uses real mouse events (works on Angular/React SPAs and CSP-strict sites like Google, Stripe). A text selector looks inside an open modal dialog first (a <dialog> opened with showModal, or a visible element with aria-modal="true"), exact text before partial. If the mouse would not reach the target at its center - another element lies in front of it (an overlay, a dialog's backdrop), the target does not take clicks there (pointer-events, visibility, clipping), or a different control inside it would get the click (a delete button on a card; for a container that is not clickable itself, its own plain link with a real address (a hash route with a path, such as #/orders, counts; a fragment with no target on the page does not) does not count, nor does a web component's own button that covers it, nor a box with a checkbox, switch or radio role right next to a hidden field and the only one in the nearest label around it, a label of that field (a box inside a closed shadow root cannot be seen and is not counted), which counts as that field) - the answer is ok:false with error "covered" and covered_by. Found before the mouse is sent, nothing is sent; if it comes in front afterwards, the rest of the presses and clicks are stopped (a release the page moves elsewhere is let through), and the answer also has maybe_landed: true, because the page may already have reacted. Examples: "button:text(Get started)", "text=Submit", "#my-button", "a.btn-primary"browser_fillFill a form input field with a value. Supports CSS selectors AND text-based selectors. Auto-scrolls and focuses the element. Works on CSP-strict sites via Chrome Debugger API. For date inputs use browser_set_date, for autocomplete/combobox use browser_set_combobox. With a text selector the field is clicked first, so the same check as browser_click applies: if the mouse would not reach it at its center, nothing is typed and the answer is ok:false with error "covered".browser_press_keyPress a keyboard key (Enter, Tab, Escape, ArrowDown, etc.). Useful for submitting forms, navigating dropdowns, closing dialogs. Supports modifier keys (ctrl, alt, shift, meta). While ctrl, meta or alt is held, no character is typed. With ctrl or meta (and no shift), the letters a, c, x, v, z and y are sent with Chrome's own editing command - select all, copy, cut, paste, undo, redo; z with shift is redo - including on macOS, where Chrome runs no shortcut for a synthetic key without it. The answer names the command that was sent in `shortcut`; `landed` says the page received the key, not what the command did. Paste inserts the user's real clipboard into the page, and copy and cut replace what is on it.- +28 more on the listing page
Browserctl MCP tools 25
browser_load_tools[EXTEND] Load a profile of tools that are not currently visible: network (capture every request, read response bodies, export a HAR, wait for network idle), cookies (read, set, delete), storage (localStorage, sessionStorage), console (console messages and page errors), cdp (raw CDP, coordinate input, audit), record (record and replay), tabs (windows, groups, visibility), advanced (a11y tree, PDF, history, element screenshots), system (a shell command on the bridge host, not the page), or all. browser_action reaches any single action without loading its profile; this is for when you want the tools themselves.browser_list_available_tools[EXTEND] Every capability this server has, loaded or not, with its parameters. Use it before assuming something is missing; browser_load_tools turns any of it on.browser_action[EXTEND] Dispatch any protocol action by name, including ones whose tool is not loaded: browser_action({action, params}). Called with no arguments it prints the catalogue — every action the bridge will dispatch. Check it before concluding something is impossible.browser_status[SESSION] Whether the bridge is reachable, the daemon's state, and whether the Chrome extension is connected. Call it when a command failed for a reason that sounds like infrastructure rather than the page; browser_start brings the daemon back.browser_snapshot[READ] A text census of the page's controls: one line per element with a stable 'ref' to act on, in reading order. Start here to see what is on a page. scope: 'viewport' (default) or 'all' — every element currently in the DOM, worth it whenever a COUNT or a COMPLETE list is the answer. 'all' is not everything the page can show: feeds and virtualised lists keep most rows out of the DOM until something is clicked, and 'hiddenContent' names the control that loads them. The census hoists key inputs and search fields to the top, and dense repetitive runs are folded; each folded control comes back in 'folded' as a line '@ref "text" -> href'. What it withheld comes back as data: window/next (paging), offscreenCount, foldedCount, duplicateCount, structure (a ref per region), pageState.openDialogs, hiddenContent. It does NOT carry pixel geometry, class names or attributes: browser_extract({selector, fields}) returns those for every match, and is cheaper for a list's data.browser_read_page[READ] The accessibility tree as indented text — which control sits inside which group, form or region — with a ref on each interactive element. Structure, not prose. Reach for browser_snapshot first unless nesting is the question, and for browser_get_property when what you want is a region's text. Narrow with target (one subtree) and mode: 'all' (include non-interactive nodes); depth defaults to 60 because a React SPA nests 25-45 levels deep, and a clipped walk reports 'depthClipped' rather than looking like an empty page.browser_find[READ] Find things on the page and get a ref back for each. in: 'controls' (default) matches interactive elements by accessible name, text, placeholder, aria-label or title, and 'matchedBy' says which of those hit; in: 'text' searches the page's prose instead and each match carries 'nearestInteractive'. SCOPE differs by index and every result names it in 'searchedScope'. Controls reads the WHOLE page — top frame, open Shadow DOM, iframes — on screen or not, so its count can differ from browser_snapshot's (viewport by default); a sub-frame match carries a ref like 'f3:ref_5', passed back verbatim. Text reads the top frame and Shadow DOM, NOT iframes. Takes a CSS 'selector' instead of 'query' when that is what you have — the cheap way to get a ref for something that just appeared, without another browser_snapshot. A zero-match answers with 'nearest': labels that differ only by case or diacritics, with their refs, plus the page's own vocabulary — so a one-character transcription error costs one call, not four. A label cut short carries 'truncatedBy' and the read that returns the rest.browser_click[ACT] Click an element by 'target' (ref '@ref_1', CSS selector, visible text, or snapshot index). Waits for the DOM to settle and reports what changed in 'effect' and 'resolved'. If the click navigates, the response reports urlChanged instead of failing on a stale ref.browser_type[ACT] Put text into any editable target — input, textarea, contenteditable, or rich-text editor. method: 'set' (default, native setters so React and Vue see it) | 'type' | 'paste' (for large payloads and AST editors). On an autocomplete field, effect.suggestions lists the options the text opened, each with a ref to click: picking one is what commits the value. For a <select> dropdown, use browser_select_option.browser_fill_form[ACT] Set several form controls in one round-trip: text fields, checkboxes, radios and <select>s. On failure, stops and reports the failed index and which fields were already written. Optionally clicks submitTarget afterwards.browser_select_option[ACT] Select one or more options in a <select> element. Values are matched by value first, then by visible label.browser_scroll[ACT] Scroll the page, or a specific container when the window itself does not move — a drawer, a table, an overflow:auto div. A container that cannot scroll is refused with the reason, rather than reported as a scroll that did nothing. To see what is below without moving, browser_snapshot({scope: 'all'}).- +13 more on the listing page
Ready-to-paste client configs
Paste either block (or both) into your client config, e.g. claude_desktop_config.json or ~/.cursor/mcp.json. We only show a config when the install command is confirmed.
Browser MCP
{
"mcpServers": {
"browser-mcp": {
"command": "npx",
"args": [
"-y",
"@agent360/browser-mcp"
]
}
}
}
Browserctl MCP
{
"mcpServers": {
"browserctl-mcp": {
"command": "npx",
"args": [
"-y",
"-p",
"browserctl-mcp",
"browserctl-mcp"
]
}
}
}
Frequently asked questions
Browser MCP is categorized under Browser Automation and uses a local stdio subprocess. Browserctl MCP belongs to Browser Automation and uses a local stdio subprocess. Pick Browser MCP for browser automation work and Browserctl MCP for browser automation.
Copy the mcpServers block from the config section on this page into your client's config (for example claude_desktop_config.json or ~/.cursor/mcp.json), then fully restart the client.
Browser MCP is listed as free. Auth: no auth required. Browserctl MCP is listed as free. Auth: no auth required. If a server needs environment variables such as API keys, define them under the env key in your client's MCP config.
Yes. MCP clients support multiple servers: add browser-mcp and browserctl-mcp as separate keys in the same mcpServers object, and your assistant will call tools from either one.
On AllMCPs, Browser MCP has a quality signal of 69/100 (Great) with 7 views, 0 installs and 58 GitHub stars. Browserctl MCP has 63/100 (Good) with 1 views, 0 installs and 0 GitHub stars.
