Control a real Chrome browser to complete any task: fill forms, extract data, book flights.
Camoufox is installed from its publisher's own source and answers where it runs, so this marketplace is not in the path of a single call. There is no address here to send one to, and a panel that pretended otherwise would be showing you an answer we made up. Install it and call it from your own client — the Installation tab has the entry for each one.
What it does
Anti-detection browser automation with Camoufox - stealth Firefox for web scraping
Quickstart
# 1 — add it to Claude Code
claude mcp add --scope user baixianger-camoufox-mcp '--' npx -y camoufox-mcp
# 2 — ask your agent for something these tools do
# list_pages, new_page, select_page and 39 more
Camoufox costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.
Collected from a public index. Nobody has claimed this account, so nothing here was written by its author — claim it if it is yours.
Where are you running it?
Every route below installs the same thing and ends at the same approval screen. This one runs on your machine: your client starts Camoufox as a process under your own user, with your files and your network, so the tool surface below is what it can do to your computer rather than to a server somewhere else. Its scan result is on the Security tab — read the surface before you approve it.
This is a public server: you run it yourself and this marketplace is not in the path. You need Node.js — the current LTS (24) is the safe choice; npx comes with it. Claude Code registers it in one command; --scope user makes it available in every project. The '--' is in quotes so that PowerShell passes it on.
claude mcp add --scope user baixianger-camoufox-mcp '--' npx -y camoufox-mcpReconnect, or start a new session, and the tools appear in the model’s tool list.
One config entry your client uses to start the process locally. A local server runs with your file system and your network, which is why it is priced without metering.
42 tools, with what each one reads, writes and reaches shown before you agree — the same list on every route above. Read the tool surface.
Tool surface
What the model actually sees. Descriptions are diffed on every release — see version history.
List all open browser pages/tabs
No parameter schema on file.
Create a new browser page/tab, optionally navigating to a URL. Specify sessionId to create in a specific session.
No parameter schema on file.
Set a page as the active page for subsequent operations
No parameter schema on file.
Close a browser page/tab
No parameter schema on file.
Navigate to a URL or go back/forward/reload
No parameter schema on file.
Take a text snapshot of the page with interactive element UIDs for targeting
No parameter schema on file.
Take a screenshot of the page or a specific element
No parameter schema on file.
Execute JavaScript in the page
No parameter schema on file.
Execute JavaScript in an ISOLATED context (sandboxed, separate from page scripts). USE THIS TOOL WHEN YOU NEED TO: - Make async operations (fetch, setTimeout, Promises) - Query/manipulate the DOM without affecting page state - Run code that shouldn
No parameter schema on file.
Extract Next.js/Nuxt.js hydration state from the page
No parameter schema on file.
Click an element by its UID from snapshot
No parameter schema on file.
Hover over an element by its UID
No parameter schema on file.
Fill a form field (input, textarea, select) by its UID
No parameter schema on file.
Fill multiple form fields at once
No parameter schema on file.
Press a keyboard key or combination
No parameter schema on file.
Drag an element to another element
No parameter schema on file.
Wait for text or selector to appear on the page
No parameter schema on file.
Wait for page navigation to complete
No parameter schema on file.
Dismiss cookie consent popups and overlays
No parameter schema on file.
Extract text content from the page
No parameter schema on file.
Reload settings from config file
No parameter schema on file.
Get all config file paths (user config, package default, active)
No parameter schema on file.
Initialize user config directory (~/.config/camoufox-mcp) with default settings
No parameter schema on file.
Start recording user interactions (clicks, scrolls, inputs)
No parameter schema on file.
Stop recording and return captured actions
No parameter schema on file.
Get current recording status
No parameter schema on file.
Start capturing network requests
No parameter schema on file.
Stop capturing network requests
No parameter schema on file.
Get captured network request logs
No parameter schema on file.
Clear captured network logs
No parameter schema on file.
Intercept and block/modify network requests
No parameter schema on file.
Remove network intercept rules
No parameter schema on file.
Start capturing browser console messages
No parameter schema on file.
Stop capturing console messages
No parameter schema on file.
Get captured console messages
No parameter schema on file.
Clear captured console logs
No parameter schema on file.
Inject browser context (cookies + localStorage) from a JSON file into a session or page\
No parameter schema on file.
Create a new isolated browser session with its own cookies and localStorage. Optionally pre-load context from a JSON file.
No parameter schema on file.
List all browser sessions with their page counts and metadata
No parameter schema on file.
Export a session\
No parameter schema on file.
Close a browser session and all its pages
No parameter schema on file.
List available context files registered in settings. Use create_session with a contextPath to load one on demand.
No parameter schema on file.
- Every tool, no call limit
- No card, and mcprush charges nothing for it
- Source published under MIT
- Runs on your machine — nothing of it reaches our gateway
- Nothing to cap, because mcprush bills nothing
What counts against your monthly calls
| Tool | Unit | Calls used | Out of the allowance |
|---|
Nothing here is billable: mcprush charges nothing to install Camoufox or to call it, at any volume.
Two independent axes, because powerful and malicious are different questions. The grade is threat only. The capability level is blast radius, and it is never a penalty on the grade — it is priced as one subtract-only term in the score, where you can see it.
| Term | Level | What it prices | Points |
|---|---|---|---|
| capability-exposure | high | capability blast radius (high) — client exposure if the model is manipulated | −6 |
| verification-discount | repo | publisher verification (public source) — no provenance, but the source is public and inspectable | −1 |
| coverage-honesty | source | inspection depth (source) — how much of the target the scan could see | −0 |
What the scan could actually read
A grade is only as meaningful as its coverage, so the scanner publishes its own depth before it publishes its result.
Tools were statically extracted from the published source (42 recovered), not enumerated from a running server. Tool-poisoning, Unicode-smuggling, capability and toxic-flow analysis ran on this inferred surface, but a mis-parsed registration could be missed or mis-attributed, so tool-derived findings are capped below “confirmed”. To grade the real runtime surface, scan the running server: --command "npx -y <package>".
Capability — what it could do if the model were manipulated
Tags derived from each tool’s schema and the implementation, not from what the tool calls itself. high is the level these add up to.
| Tool | Capability tags | Why the tag was assigned |
|---|---|---|
| list_pages | no tags | |
| new_page | no tags | |
| select_page | no tags | |
| close_page | no tags | |
| navigate_page | no tags | |
| take_snapshot | no tags | |
| take_screenshot | no tags | |
| evaluate_mainworld | code-exec | code-exec: parameter "script" |
| evaluate_isolated | code-exec | code-exec: parameter "script" |
| get_framework_state | no tags | |
| click | no tags | |
| hover | no tags | |
| fill | no tags | |
| fill_form | no tags | |
| press_key | no tags | |
| drag | no tags | |
| wait_for | no tags | |
| wait_for_navigation | no tags | |
| dismiss_popups | no tags | |
| get_page_text | untrusted-input | untrusted-input: keyword "get_page" |
| reload_settings | no tags | |
| get_config_paths | no tags | |
| init_user_config | no tags | |
| start_recording | no tags | |
| stop_recording | no tags | |
| get_recording_status | no tags | |
| start_network_capture | no tags | |
| stop_network_capture | no tags | |
| get_network_logs | no tags | |
| clear_network_logs | no tags | |
| intercept_network | no tags | |
| remove_intercept | no tags | |
| start_console_capture | no tags | |
| stop_console_capture | no tags | |
| get_console_logs | no tags | |
| clear_console_logs | no tags | |
| inject_context | no tags | |
| create_session | no tags | |
| list_sessions | no tags | |
| save_session | no tags | |
| close_session | no tags | |
| list_context_files | no tags |
Toxic-flow graph
The lethal trifecta, checked as a graph rather than as a checklist: untrusted input, a sensitive source and an external sink have to meet before there is a path worth worrying about.
Supply chain and provenance
This is the first scan of this surface here, so there is nothing yet to compare it against.
Every result on this tab comes from one deterministic pass over the published package — offline, rule by rule, and auditable line by line above. Same methodology version, same bytes, same score.
Release history
Always latest — the install command names no version, so it installs the newest release when it runs; the current release on file is 1.0.0. Pin a release below and the command asks for exactly that one.
No release note was published with this version.
One review per account, from a signed-in account that does not publish this listing. Nothing else is asked — you can say what you think before you install it. Publishers can reply once per review, and a review can be edited or deleted by whoever wrote it.
Any signed-in account that does not publish Camoufox can review it, once — before installing it or after. A review from an account that has installed it is marked as one.
Nobody has reviewed this listing. The rating on the card is the mean of the reviews written here and nothing else, so there is no rating until somebody writes the first — which takes a signed-in account that does not publish it, and nothing else.