chrome-mcp server

Read from the package source without running it. The installed server may list more.

• auth_check

  Is the tab sitting on a sign-in wall? Reads the page (URL, title, password fields, sign-in controls) and returns { authRequired, confidence, signals }. Use it after a navigate, or whenever a step fails unexpectedly, to tell "the session expired" apart from "the agent got lost". Pass failOnAuthWall:true to get an [AUTH_REQUIRED] error instead of a verdict, so a harness can bucket the run as an auth

• back

  Go back in history.

• batchwrite action

  Run multiple tool calls in one request — parallel (default) or serial. Each op is { tool, args } and goes through the same policy gate, rate limit, and error handling as a direct call. In parallel mode, tab-scoped ops MUST pass an explicit tabId (the active-tab default is unsafe under concurrency). Use to drive several tabs at once (e.g. open tabs, then batch get_text across them). Cannot be neste

• chrome_status

  Report backend/session status: paired browser profiles and how each was named, and flags for switched-off capabilities.

• click

  Click an element. Target by selector, a snapshot ref, or role+name (e.g. role:"button", name:"Sign in") - the locator needs no snapshot first. trusted=true uses real OS-level input.

• console_logs

  Console output and uncaught errors recorded on the page (requires --enable-observers). This is how you find out WHY a page misbehaved rather than only what it looks like afterwards. Pass sinceSeq to poll for what is new, clear:true to drain.

• dialogs

  Native dialogs (alert/confirm/prompt/beforeunload) the page raised, and how they were answered (requires --enable-observers). With observers on, dialogs are intercepted rather than left to block the renderer - which is what otherwise turns a click that opens a confirm() into a mystery TIMEOUT. Set policy:'accept' to answer confirms with true.

• download_file

  Download a file by URL or from a link element.

• eval

  Evaluate JavaScript in the page (disabled in safe-mode).

• extract_links

  Extract anchors from the page or a subtree. dedupe=true collapses links sharing an href (nav/footer noise); limit caps the count.

• fill_form

  Fill multiple fields (keyed by selector) and optionally submit.

• forward

  Go forward in history.

• frames_list

  List the tab's frames (the top document plus every iframe the extension can reach), with each frame's id and URL. Use it when a selector that should match does not: the element is probably in one of these frames, and you can then pass frameId (or allFrames:true) to act inside it.

• get_cookies

  Read cookies visible to the tab's URL (or a given url).

• get_html

  Get HTML of the page or an element. Output is capped (see maxBytes) and cut at a tag boundary; narrow it with `selector` rather than raising the cap when you can. Password field values are always blanked.

• get_text

  Get visible text of the page or an element.

• hover

  Hover over an element.

• navigate

  Navigate a tab to a URL, REPLACING its current page. Acts on the active tab unless tabId is given — to open a site without losing the current page, use `tab_new` instead.

• network_log

  Requests the page made - fetch and XMLHttpRequest with method, URL, status and duration (requires --enable-observers). Set includeResources:true to also list scripts/images/styles from Resource Timing (those carry timing and size but no status). Does not include the document request or headers.

• press

  Press a key (with optional modifiers).

• print_pdf

  Render the page to PDF through Chrome's own print pipeline and save it to the task's results/ dir. Returns the path and size, not the bytes - a PDF is not something to spend context on.

• profile_rename

  Rename an automatically named browser profile (e.g. "profile-2" -> "work"). The name sticks across restarts; its saved artifacts move with it.

• profile_use

  Switch the active browser profile (identity). Subsequent downloads, results, screenshots, and the action log are stored under profiles/<name>/. Resets the active task to "default" unless you then call task_new.

• read_as_markdown

  Read the page (or subtree) as readable markdown.

• reload

  Reload the active (or given) tab.

• screenshot

  Capture a screenshot (page or element). Default is JPEG (quality 70) at CSS-pixel size, which is several times smaller than PNG and reads fine. Pass format:"png" for lossless, quality 1-100 for JPEG, scale 2 for device pixels on a Retina display or 0.5 to shrink.

• scroll

  Scroll the page or to an element.

• select_option

  Select option(s) of a <select> by value or visible label.

• snapshot

  Accessibility snapshot: interactive elements with refs to target by `ref` (more reliable than guessing CSS selectors). Pass diff:true to get only what changed since the last snapshot of this tab - far cheaper in a click/read loop. Password fields appear as secret:true with no value.

• storage

  Read/write localStorage (or sessionStorage). op: get|set|remove|clear.

• tab_close

  Close a tab by tabId.

• tab_new

  Open a NEW tab (optionally at a URL) and focus it. Prefer this over `navigate` when the user says "open"/"go to" a site — `navigate` REPLACES the current tab. Pass active:false to open in the background (used by parallel batches).

• tab_select

  Make a tab active by tabId.

• tabs_list

  List open browser tabs.

• task_new

  Start a new task (run) under the active profile. Creates profiles/<profile>/tasks/<name>/ with downloads/, results/, screenshots/ and makes it the active task so all captured artifacts land there.

• task_status

  Report the active profile/task and the folder paths where this run's artifacts are stored.

• tasks_list

  List every task across all profiles under the data dir, with sizes and download counts.

• type

  Type text into an element (target by selector, ref, or role+name). trusted=true sends real keystrokes (works on React/Vue controlled inputs).

• upload_filewrite action

  Set local file(s) on a file <input> (target by selector or ref) — uploads without the OS dialog. Requires --enable-uploads. `files` are absolute local paths.

• wait_for

  Wait for a selector or text to appear/disappear.

• Code scan91 source files scanned20/25
• –Live reliabilityno gateway calls yet and no remote to proben/a
• –Tool poisoningtools not inspected (local package is not executed); not countedn/a
• Auth qualitylocal package, no credentials required12/15
• Maintenancelast push 0 days ago15/15
• Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10

chrome-mcp: common questions

Is chrome-mcp server safe?

Yes, by our scan: it is graded A (85/100). Read the chrome-mcp safety report

How do I install chrome-mcp?

It runs on your machine. Copy the Claude Code, Claude Desktop or Cursor config from the install section.

Does chrome-mcp need an API key?

Not as far as the registry entry and our scan can tell: no credentials are declared or required.

Is chrome-mcp maintained?

The last commit was in the last day (2026-09-23). The latest release is v0.9.11.

What can I use instead of chrome-mcp?

Servers from other publishers that do the same job: DOMShell MCP server, Smart Connections MCP server and Ridealong MCP server. Compare all chrome-mcp alternatives.