Skip to content

scrapeunblocker-mcp-remote

com.scrapeunblocker/scrapeunblocker-mcp-remotescrapeunblocker.comstreamable-httpv0.14.3Scanned

Web search and content extraction for AI: Google results, any page as Markdown or JSON

25 tools: 2 can delete or cancel things, 5 change things and 18 only read.

Tools: 2 delete, 0 irreversible, 5 write, 18 read.
  • Reads private data · 8
  • Reads text others write · 25
  • Can send data out · 7
  • 7 findings

Tools listed without sign-in. Remote posture

Anlyon alternative

7 of these tools change something. Run those as governed actions.

Declare what an action changes, and Anlyon puts it under a shared limit, binds every approval to the exact request, and grades the outcome. It works on any HTTPS API you can describe as an action.

Anlyon Radar does not run, proxy or govern this MCP server. A tool your own code calls directly is not routed through Anlyon and is not governed by it.

Import an OpenAPI spec →

Its sign-in does not let Anlyon set up an OAuth client on its own, so there is no link to it here.

What the scrapeunblocker-mcp-remote can do

The scrapeunblocker-mcp-remote is published by scrapeunblocker.com, a domain the MCP Registry verified. It has 25 tools. It can delete or cancel things (2 tools: browse_page and capture_xhr) and create or change data (5 tools: google_search, run_plugin, google_shopping, google_local and 1 more). 18 tools only read.

8 tools read private data (fetch_html, browse_page, fetch_parsed and 5 more); 25 tools bring in text other people can write (fetch_html, browse_page, capture_xhr and 22 more); 7 tools can send data out of your system (fetch_html, fetch_parsed, list_elements and 4 more). On its own it has all three parts of a prompt-injection exfiltration path: private data, text others write, and a way out.

The fixed checks matched 7 findings (annotation-conflict and long-description). Each one comes with the text it matched.

Its remote endpoint lists its tools without sign-in, and its tools changed 10 times since Anlyon Radar first saw them.

Last scanned Nov 19, 2026, version 0.14.3.

Tools

Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.

25 tools

Deletes or cancels 2Claude Code: deny
  • browse_pageLoad a web page in a real browser through the ScrapeUnblocker API (https://docs.scrapeunblocker.com), run ordered browser actions on it (click, type into a search box, choose an op…Reads private dataReads text others write1 finding

    Why it is in this group

    • The name "browse_page" says "browse": a word for reading.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • private-dataThe description's first sentence says "doc": it reads documents.
    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe name "browse_page" says "browse": it reads web pages, which anyone can write.
    • untrusted-inputThe description's first sentence says "scrape": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.

    Findings

    • Warn

      The server marks "browse_page" destructive (destructiveHint: true), and its name says "browse", a word for reading.

      Matched text

      annotations: {"readOnlyHint":false,"destructiveHint":true,"idempotentHint":false,"openWorldHint":true}; The name "browse_page" says "browse": a word for reading.

      What this check looks for
    Description, as the server wrote it
    Load a web page in a real browser through the ScrapeUnblocker API (https://docs.scrapeunblocker.com), run ordered browser actions on it (click, type into a search box, choose an option, press a key, scroll, wait), then return the resulting HTML. Steps act on the website and are NOT idempotent - they run once per call and are not retried. A failed step returns which step failed plus the page HTML at that point. Billed as one request.

    Annotations the server declares

    • title: Interact with a page, then capture it
    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
  • capture_xhrLoad a web page in a real browser through the ScrapeUnblocker API (https://docs.scrapeunblocker.com) and return the JSON its own API calls received: every XHR/fetch response whose …Reads text others writeNo findings

    Why it is in this group

    • The name "capture_xhr" says "capture": a word that usually means money moves, which cannot be taken back.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "scrape": it reads web pages, which anyone can write.
    Description, as the server wrote it
    Load a web page in a real browser through the ScrapeUnblocker API (https://docs.scrapeunblocker.com) and return the JSON its own API calls received: every XHR/fetch response whose URL contains `capture_xhr` (a fragment such as '/api/search'), in order, with status and body. Use it when a site renders from a JSON API that refuses direct requests: the page makes the call, so it succeeds. Optional `steps` make the page fire the call (a search, 'load more'); steps act on the website and run once. Up to 20 responses; the page HTML is not returned. Billed as one request; when the site refuses every matching call (403/429) the call fails and is free.

    Annotations the server declares

    • title: Capture a page's own API calls (XHR/fetch)
    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
Changes things 5Claude Code: ask
  • google_imagesRuns the Google Images vertical (`udm=2`) for `q` and returns the image results as JSON.Reads text others write1 finding

    Why it is in this group

    • The description's first verb is "run": a word for a change.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.

    Findings

    • Warn

      The server marks "google_images" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"openWorldHint":true}; The description's first verb is "run": a word for a change.

      What this check looks for
    Description, as the server wrote it
    Runs the Google Images vertical (`udm=2`) for `q` and returns the image results as JSON. Every result carries `imageUrl` (the full-resolution origin image, not a Google thumbnail; for an image found inside a document such as a PDF, which has no URL of its own, it is Google's thumbnail and `sourcePageUrl` is the document), `sourceDomain` and `googleImagesUrl` (the Google Images viewer link, i.e. the `/imgres?...` URL a browser copies from the result tile), plus `sourcePageUrl`, `title`, `sourceName`, `thumbnailUrl`, `width`, `height` and `fileSize` where available. Set `proxy_country` to target a market: the Google market (`gl`) follows it automatically, and `gl` is only an optional override. `pages` fetches up to 5 result pages of ~100 each; every page fetched is billed as one request. `time_range` (past hour/day/week/month/year) or `date_from`/`date_to` limit the results to images Google found in that period, so a repeated scan returns only what is new. Returns JSON. Billed like an API request; listing plugins bill one unit per result page they fetch, so keep max_results modest.

    Annotations the server declares

    • title: Scrape Google Images results
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • google_localRuns a Google Local (`tbm=lcl`) search for `keyword` and returns up to ~20 businesses as JSON: `name`, `rating`, `reviews`, `price`, `category`, `address`, `phone`, `website`, `hou…Reads text others write1 finding

    Why it is in this group

    • The description's first verb is "run": a word for a change.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.

    Findings

    • Warn

      The server marks "google_local" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"openWorldHint":true}; The description's first verb is "run": a word for a change.

      What this check looks for
    Description, as the server wrote it
    Runs a Google Local (`tbm=lcl`) search for `keyword` and returns up to ~20 businesses as JSON: `name`, `rating`, `reviews`, `price`, `category`, `address`, `phone`, `website`, `hours`, and a top `review_snippet`. `phone` and `website` are null when Google does not render them for that result - service businesses (plumbers, contractors) almost always carry both. Local results are location-sensitive - set `proxy_country` (and optionally `gl`/`hl`) to target a market. Returns JSON. Billed like an API request; listing plugins bill one unit per result page they fetch, so keep max_results modest.

    Annotations the server declares

    • title: Scrape Google Local business listings
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • google_shoppingRuns the Google Shopping vertical (`udm=28`) for `q` and returns the product cards as JSON: `title`, `price`, `currency`, `currencyCode`, `originalPrice`, `discount`, `merchant`, `…Reads text others write1 finding

    Why it is in this group

    • The description's first verb is "run": a word for a change.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.

    Findings

    • Warn

      The server marks "google_shopping" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"openWorldHint":true}; The description's first verb is "run": a word for a change.

      What this check looks for
    Description, as the server wrote it
    Runs the Google Shopping vertical (`udm=28`) for `q` and returns the product cards as JSON: `title`, `price`, `currency`, `currencyCode`, `originalPrice`, `discount`, `merchant`, `delivery`, `rating`, `reviewCount`, `image` and Google's own ids for the product (`productId`, `catalogId`, `offerId` - stable across searches, for tracking one product). Set `proxy_country`/`gl` to target a market (prices/merchants are region-specific) and `min_price`/`max_price` to filter. One render returns the first page of products (v1). Note: Google Shopping exposes no per-product URL, so `url` is omitted. `image` is the product's thumbnail URL. Returns JSON. Billed like an API request; listing plugins bill one unit per result page they fetch, so keep max_results modest.

    Annotations the server declares

    • title: Scrape Google Shopping product listings
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • run_pluginRun one ScrapeUnblocker site plugin and return its JSON - e.g.Reads text others write1 finding

    Why it is in this group

    • The name "run_plugin" says "run": a word for a change.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "scrape": it reads web pages, which anyone can write.

    Findings

    • Warn

      The server marks "run_plugin" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"openWorldHint":true}; The name "run_plugin" says "run": a word for a change.

      What this check looks for
    Description, as the server wrote it
    Run one ScrapeUnblocker site plugin and return its JSON - e.g. a Target or Walmart search, a Zillow listing search, Google Trends for some keywords, an app's App Store reviews. `params` uses the plugin's own parameter names (snake_case); a list value is fine where a parameter takes several. Each run is billed like an API request; listing plugins bill one unit per result page they fetch, so max_results sets the cost.

    Annotations the server declares

    • title: Run a site plugin
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
Only reads 18Show 18 read-only tools
  • amazon_productFetches a single Amazon product and returns it as JSON: `title`, `brand` and `brandStoreUrl`, `parentAsin`, `manufacturer`, `modelNumber`, `partNumber`, `productCodes` (UPC, EAN, G…Reads private dataReads text others writeCan send data out1 finding

    Why it is in this group

    • The description's first verb is "fetch": a word for reading.

    What it can reach

    • private-dataThe description's first sentence says "fetch" and "document": it reads documents.
    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "fetch" and "url": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Findings

    • Info

      The description is 2,036 characters long, over the 2,000 this check looks for.

      Matched text

      description ends: "...ing plugins bill one unit per result page they fetch, so keep max_results modest."

      What this check looks for
    Description, as the server wrote it
    Fetches a single Amazon product and returns it as JSON: `title`, `brand` and `brandStoreUrl`, `parentAsin`, `manufacturer`, `modelNumber`, `partNumber`, `productCodes` (UPC, EAN, GTIN, ISBN), `condition`, the `amazonsChoice`, `bestSellerBadge`, `deal` and `coupon` markers, `price` + `currency`, `listPrice`, `savingsPercent` and `savingsAmount`, `unitPrice`, `subscribeAndSavePrice`, `shippingPrice`, `recentSales`, `availability` (+ a boolean `inStock`), `maxOrderQuantity`, `rating`, `reviewCount`, `ratingDistribution` (percent per star), `reviewSummary`, `seller`, `sellerUrl`, `sellerId`, `shipsFrom`, `fulfilledByAmazon`, `prime`, `delivery`, `returnPolicy`, `offersCount` and `offersFromPrice`, `bestSellersRank`, `features` (bullet points), `categories` (breadcrumb), `description`, `fromTheManufacturer` and `fromTheBrand` (text and images), `importantInformation`, `ingredients`, `whatsInTheBox`, `documents`, `warrantyAndSupport`, `productDetails`, `productOverview`, `dimensions`, `weight`, `dateFirstAvailable`, `variantDimensions` and `variants` (each with its own ASIN and options), `sustainability`, `protectionPlans`, `newerModel`, `sponsoredProducts`, `relatedProducts`, `images` (the whole gallery) and `videos`. A field the page does not show is null or empty. Identify the product with either `asin` (plus `marketplace`) or a full product `url`. Other sellers' full offers are not on the product page: only their count and lowest price are returned here, `/marketplace/amazon-offers` lists them. `prime` and `delivery` are what a visitor who is not signed in sees. Amazon prices each page in the currency of the delivery location it infers from the visitor's IP address, so `proxy_country` defaults to the marketplace's home country (amazon.com -> US) and the call is routed through proxies in that country. The country of the proxy that served the request is reported as `exitCountry`. Returns JSON. Billed like an API request; listing plugins bill one unit per result page they fetch, so keep max_results modest.

    Annotations the server declares

    • title: Scrape an Amazon product page
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • amazon_reviewsReturns the review data an Amazon product page shows to a signed-out visitor, as JSON: the overall `rating`, total `reviewCount` (global ratings), the 5..1-star `histogram` (percen…Reads text others writeCan send data outNo findings

    Why it is in this group

    • The description's first verb is "return": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe name "amazon_reviews" says "review": it reads comments, which other people can write.
    • untrusted-inputThe description's first sentence says "return" and "review": it reads comments, which other people can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Returns the review data an Amazon product page shows to a signed-out visitor, as JSON: the overall `rating`, total `reviewCount` (global ratings), the 5..1-star `histogram` (percent per star), the AI `customersSay` summary (overall text plus per-aspect `sentiment`, `mentions`, `positive`, `negative`, `summary` and the quoted review `excerpts` Amazon shows for the aspect) and the top `reviews` (each with `id`, `url`, `author`, `rating`, `title`, `body`, `date`, `country`, `fromOtherCountry`, `verified`, `helpfulVotes`, `variant` and `images`). Identify the product with `asin` (plus `marketplace`) or a product `url`. Scope: Amazon only shows its full review list to signed-in customers, so this endpoint returns the public slice on the product page - typically 5 to 13 top reviews, local ones first and then reviews from other countries - not a complete review export. When Amazon hides the top reviews behind a sign-in prompt (common on amazon.com), `reviewsGated` is `true` and `reviews` is empty; the rating, histogram and customers-say data, including its review excerpts, are still returned. Every field is always present and is `null` when the page does not show it. `proxy_country` defaults to the marketplace's home country; the country of the proxy that served the request is reported as `exitCountry`. Returns JSON. Billed like an API request; listing plugins bill one unit per result page they fetch, so keep max_results modest.

    Annotations the server declares

    • title: Scrape the reviews on an Amazon product page
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • booking_hotelsSearches Booking.com for `location` and returns the properties as JSON: `name`, `price` (with `checkin`/`checkout`), `reviewScore`, `address`, coordinates, `image` and a hotel `url…Reads text others writeNo findings

    Why it is in this group

    • The description's first verb is "search": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "search" and "review": it reads comments, which other people can write.
    Description, as the server wrote it
    Searches Booking.com for `location` and returns the properties as JSON: `name`, `price` (with `checkin`/`checkout`), `reviewScore`, `address`, coordinates, `image` and a hotel `url`. Set `adults`, `children`, `rooms` and `currency`, and collect up to `max_results` across pages. Prices only appear when check-in/check-out dates are given. Returns JSON. Billed like an API request; listing plugins bill one unit per result page they fetch, so keep max_results modest.

    Annotations the server declares

    • title: Scrape Booking.com hotels
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • fetch_htmlFetch the fully rendered HTML of any public web page through the ScrapeUnblocker API (https://docs.scrapeunblocker.com).Reads private dataReads text others writeCan send data outNo findings

    Why it is in this group

    • The name "fetch_html" says "fetch": a word for reading.

    What it can reach

    • private-dataThe description's first sentence says "fetch" and "doc": it reads documents.
    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe name "fetch_html" says "fetch" and "html": it reads web pages, which anyone can write.
    • untrusted-inputThe description's first sentence says "scrape": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Fetch the fully rendered HTML of any public web page through the ScrapeUnblocker API (https://docs.scrapeunblocker.com). The page loads in a real browser, so JavaScript-rendered pages come back complete. Use it when you need the page markup, or when a plain HTTP fetch returns 403/429 or an empty page. Returns raw HTML.

    Annotations the server declares

    • title: Fetch page HTML
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • fetch_markdownFetch a web page in a real browser through the ScrapeUnblocker API (https://docs.scrapeunblocker.com) and return it as clean Markdown: the main content with absolute links, without…Reads private dataReads text others writeCan send data outNo findings

    Why it is in this group

    • The name "fetch_markdown" says "fetch": a word for reading.

    What it can reach

    • private-dataThe description's first sentence says "fetch" and "doc": it reads documents.
    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "scrape": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Fetch a web page in a real browser through the ScrapeUnblocker API (https://docs.scrapeunblocker.com) and return it as clean Markdown: the main content with absolute links, without scripts, styles, navigation or footer. Use it to read, quote or summarise a page; it is usually 2-25% the size of the HTML. Works on JavaScript-rendered pages and on pages that return 403/429 to a plain HTTP fetch. Billed as one request.

    Annotations the server declares

    • title: Fetch page as Markdown
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • fetch_parsedFetch a web page through the ScrapeUnblocker API (https://docs.scrapeunblocker.com) and return AI-parsed structured JSON instead of raw HTML (product details, article content, list…Reads private dataReads text others writeCan send data outNo findings

    Why it is in this group

    • The name "fetch_parsed" says "fetch": a word for reading.

    What it can reach

    • private-dataThe description's first sentence says "fetch" and "doc": it reads documents.
    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "scrape": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Fetch a web page through the ScrapeUnblocker API (https://docs.scrapeunblocker.com) and return AI-parsed structured JSON instead of raw HTML (product details, article content, listings). If the page holds no structured data, the result says so and returns the rendered HTML instead (billed like fetch_html).

    Annotations the server declares

    • title: Fetch AI-parsed page data
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • indeed_jobsSearches Indeed for `query` in `location` and returns the job listings as JSON: `title`, `company`, `location`, `salary` (with parsed `min`/`max`), `jobTypes`, `companyRating`, a `…Reads text others writeNo findings

    Why it is in this group

    • The description's first verb is "search": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "search" and "url": it reads web pages, which anyone can write.
    Description, as the server wrote it
    Searches Indeed for `query` in `location` and returns the job listings as JSON: `title`, `company`, `location`, `salary` (with parsed `min`/`max`), `jobTypes`, `companyRating`, a `snippet`, `datePosted` (YYYY-MM-DD), `sponsored` (Indeed's paid placement) and a clean job `url`. Filter with `country`, `job_type`, `days` (posting age), `remote`, sort with `sort` (`date` returns the collected jobs newest first, sponsored ones included), and collect up to `max_results` across pages. Indeed sits behind Cloudflare + PerimeterX - handled for you. Returns JSON. Billed like an API request; listing plugins bill one unit per result page they fetch, so keep max_results modest.

    Annotations the server declares

    • title: Scrape Indeed job listings
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • list_elementsFetch a page through the ScrapeUnblocker API (https://docs.scrapeunblocker.com) and return its interactive elements (buttons, inputs, selects, links, forms), each with a ready-to-u…Reads private dataReads text others writeCan send data outNo findings

    Why it is in this group

    • The name "list_elements" says "list": a word for reading.

    What it can reach

    • private-dataThe description's first sentence says "fetch" and "doc": it reads documents.
    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "scrape": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Fetch a page through the ScrapeUnblocker API (https://docs.scrapeunblocker.com) and return its interactive elements (buttons, inputs, selects, links, forms), each with a ready-to-use selector, as JSON {url, count, elements:[...]} instead of raw HTML. Use it to find the selectors a page's buttons, inputs and links answer to.

    Annotations the server declares

    • title: List interactive page elements
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • list_pluginsList ScrapeUnblocker's ready-made site plugins: scrapers for specific sites that return clean structured JSON (Amazon, Walmart, Target, eBay, Zillow, Redfin, Booking.com, Airbnb, I…Reads text others writeNo findings

    Why it is in this group

    • The name "list_plugins" says "list": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "scrape": it reads web pages, which anyone can write.
    Description, as the server wrote it
    List ScrapeUnblocker's ready-made site plugins: scrapers for specific sites that return clean structured JSON (Amazon, Walmart, Target, eBay, Zillow, Redfin, Booking.com, Airbnb, Indeed, Glassdoor, Google Trends / Shopping / Maps, YouTube, Spotify, App Store, Google Play, Trustpilot, Yelp, Ahrefs, Semrush and 100+ more). A plugin returns a site's data already parsed. Without arguments it lists every plugin with its parameters (* = required); `category` or `search` narrow the list; `name` returns one plugin's full parameter docs. Free - it calls nothing.

    Annotations the server declares

    • title: List site plugins
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • maps_place_detailsReturns the full data for one Google Maps place as JSON: `name`, `address` (+ `addressLines`), `phone`, `website`, `rating`, `categories`, `latitude`/`longitude`, `plusCode`, `hour…Reads text others writeCan send data outNo findings

    Why it is in this group

    • The description's first verb is "return": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "return" and "website": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Returns the full data for one Google Maps place as JSON: `name`, `address` (+ `addressLines`), `phone`, `website`, `rating`, `categories`, `latitude`/`longitude`, `plusCode`, `hours` + `openNow`, `description`, `photo`, `cid`, `featureId`, `googleMapsUrl` and `relatedPlaces`. Identify the place with `place_id` (ChIJ...), a Maps `url`, or a raw `feature_id` (0xHEX:0xHEX). Location-neutral - the feature id resolves the place directly; `proxy_country` defaults to a US proxy to clear Google's consent wall. Returns JSON. Billed like an API request; listing plugins bill one unit per result page they fetch, so keep max_results modest.

    Annotations the server declares

    • title: Scrape Google Maps place details
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • search_and_readSearch Google through the ScrapeUnblocker API (https://docs.scrapeunblocker.com) and read the top results in one call: returns Google's AI Overview when there is one, each of the f…Reads private dataReads text others writeNo findings

    Why it is in this group

    • The name "search_and_read" says "search": a word for reading.

    What it can reach

    • private-dataThe description's first sentence says "search" and "doc": it reads documents.
    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "scrape": it reads web pages, which anyone can write.
    Description, as the server wrote it
    Search Google through the ScrapeUnblocker API (https://docs.scrapeunblocker.com) and read the top results in one call: returns Google's AI Overview when there is one, each of the first `results` pages (default 3, max 5) as Markdown, and the other organic results as title, URL and snippet. Use it to answer a question from the live web. Billed as one request for the search plus one per page read.

    Annotations the server declares

    • title: Search Google and read the top results
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • trustpilot_reviewsReturns a company's Trustpilot profile and reviews as JSON.Reads text others writeNo findings

    Why it is in this group

    • The description's first verb is "return": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe name "trustpilot_reviews" says "review": it reads comments, which other people can write.
    • untrusted-inputThe description's first sentence says "return" and "review": it reads comments, which other people can write.
    Description, as the server wrote it
    Returns a company's Trustpilot profile and reviews as JSON. `company` is the reviewed domain (e.g. `www.nike.com`). The response has a `company` summary (`trustScore`, `totalReviews`, `ratingDistribution` 1-5, `categories`, `contact`) and a `reviews` array - each with `url`, `rating`, `title`, `text`, `verified`, `verificationLevel`, `verificationSource`, `source`, `pending`, `merged`, `author` (name, country, review count, `id`, `image`, `profileUrl`, `reviewsOfThisCompany`), `reply` (the company's response, with dates) and `date` / `experiencedDate` / `updatedDate`. Page with `page`/`pages` (20 reviews per page), and filter with `stars`, `sort` (recency/relevancy), `language`, `date`, `search`, `verified` or `replies`. Returns JSON. Billed like an API request; listing plugins bill one unit per result page they fetch, so keep max_results modest.

    Annotations the server declares

    • title: Scrape Trustpilot company reviews
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false
  • youtube_transcriptReturns a YouTube video's captions as JSON: timestamped `segments` (`start`, `dur`, `text`) and a joined `text`, plus `availableLanguages`, the picked `language`, `isAutoGenerated`…Reads text others writeNo findings

    Why it is in this group

    • The description's first verb is "return": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Returns a YouTube video's captions as JSON: timestamped `segments` (`start`, `dur`, `text`) and a joined `text`, plus `availableLanguages`, the picked `language`, `isAutoGenerated`, and the video's `title`, `author` and `lengthSeconds`. Pass a video id or any YouTube video URL; choose a caption language with `lang`. Ideal for LLM/RAG pipelines, subtitles and content analysis. Returns JSON. Billed like an API request; listing plugins bill one unit per result page they fetch, so keep max_results modest.

    Annotations the server declares

    • title: Get a YouTube video transcript
    • readOnlyHint: true
    • openWorldHint: true
    • destructiveHint: false

This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.

Exfil surface

Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.

1400 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 7 outbound tools breaks every one.

Reads private data8
Reads files, mail, messages, databases, secrets, calendars or repositories.
Reads text others write25
Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
Can send data out7
Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
deny to break every path
{
  "permissions": {
    "deny": [
      "mcp__scrapeunblocker-remote__amazon_product",
      "mcp__scrapeunblocker-remote__amazon_reviews",
      "mcp__scrapeunblocker-remote__fetch_html",
      "mcp__scrapeunblocker-remote__fetch_markdown",
      "mcp__scrapeunblocker-remote__fetch_parsed",
      "mcp__scrapeunblocker-remote__list_elements",
      "mcp__scrapeunblocker-remote__maps_place_details"
    ]
  }
}

Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.

Remote posture

What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.

Tools listed without sign-in

Looked at Nov 20, 2026, 18:46 UTC, unchanged since Nov 20, 2026. No credentials were sent and no tool was called. What is looked at.

Reachable
yes
HTTP status of an anonymous initialize
200
tools/list served without credentials
yes
Tools that change things, listed without credentials
7
Protocol version
2025-06-18
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
https://mcp.scrapeunblocker.com/.well-known/oauth-protected-resource/mcp
Authorization servers
https://login.scrapeunblocker.com/
Legacy HTTP+SSE transport
no
Dynamic client registration
no
Client ID metadata documents
no
  • HighThe whole server

    The endpoint served tools/list without credentials, and the list has 7 tools that change things (2 destructive, 5 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://mcp.scrapeunblocker.com/mcp: tools/list returned 25 tools to an anonymous client, including browse_page (destructive), capture_xhr (destructive), google_search (write), run_plugin (write), google_shopping (write) and 2 more

    What this check looks for

Posture changes

  1. Nov 20, 2026
    • Tools that change things, listed without credentials: 6 to 7

Findings

Matches of fixed checks, with the text they matched.

6 warn1 info
  • The server marks "browse_page" destructive (destructiveHint: true), and its name says "browse", a word for reading.

    Matched text

    annotations: {"readOnlyHint":false,"destructiveHint":true,"idempotentHint":false,"openWorldHint":true}; The name "browse_page" says "browse": a word for reading.

    What this check looks for
  • The server marks "google_search" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"openWorldHint":true}; The name "google_search" says "search": a word for reading.

    What this check looks for
  • The server marks "run_plugin" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"openWorldHint":true}; The name "run_plugin" says "run": a word for a change.

    What this check looks for
  • The server marks "google_shopping" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"openWorldHint":true}; The description's first verb is "run": a word for a change.

    What this check looks for
  • The server marks "google_local" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"openWorldHint":true}; The description's first verb is "run": a word for a change.

    What this check looks for
  • The server marks "google_images" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"openWorldHint":true}; The description's first verb is "run": a word for a change.

    What this check looks for
Show 1 more finding
  • The description is 2,036 characters long, over the 2,000 this check looks for.

    Matched text

    description ends: "...ing plugins bill one unit per result page they fetch, so keep max_results modest."

    What this check looks for

History

Changes to its tool definitions and to its remote endpoint, newest first.

  1. PostureRemote posture changed. Tools that change things, listed without credentials: 6 -> 7.View diff
    • Tools that change things, listed without credentials: 6 to 7
  2. Tools7 changed.~7 changedView diff
  3. Tools1 added.+1 addedView diff
  4. Tools1 changed.~1 changedView diff
  5. Tools4 changed.~4 changedView diff
  6. Tools24 changed.~24 changedView diff
  7. Tools1 changed.~1 changedView diff
  8. Tools1 changed.~1 changedView diff
  9. Tools3 changed.~3 changedView diff
  10. Tools2 changed.~2 changedView diff

Install

Pinned to version 0.14.3. Values in angle brackets are yours to fill in.

claude mcp add --transport http scrapeunblocker-remote https://mcp.scrapeunblocker.com/mcp
claude mcp add --transport http scrapeunblocker-remote https://mcp.scrapeunblocker.com/mcp

Run in a terminal.

Claude Code permissions

Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.

.claude/settings.json
{
  "permissions": {
    "allow": [
      "mcp__scrapeunblocker-remote__search_and_read",
      "mcp__scrapeunblocker-remote__list_plugins",
      "mcp__scrapeunblocker-remote__amazon_search",
      "mcp__scrapeunblocker-remote__walmart_search",
      "mcp__scrapeunblocker-remote__ebay_search",
      "mcp__scrapeunblocker-remote__youtube_search",
      "mcp__scrapeunblocker-remote__youtube_transcript",
      "mcp__scrapeunblocker-remote__zillow_search",
      "mcp__scrapeunblocker-remote__indeed_jobs",
      "mcp__scrapeunblocker-remote__trustpilot_reviews",
      "mcp__scrapeunblocker-remote__booking_hotels"
    ],
    "ask": [
      "mcp__scrapeunblocker-remote__fetch_html",
      "mcp__scrapeunblocker-remote__fetch_parsed",
      "mcp__scrapeunblocker-remote__google_search",
      "mcp__scrapeunblocker-remote__list_elements",
      "mcp__scrapeunblocker-remote__fetch_markdown",
      "mcp__scrapeunblocker-remote__run_plugin",
      "mcp__scrapeunblocker-remote__amazon_product",
      "mcp__scrapeunblocker-remote__amazon_reviews",
      "mcp__scrapeunblocker-remote__google_shopping",
      "mcp__scrapeunblocker-remote__google_local",
      "mcp__scrapeunblocker-remote__maps_place_details",
      "mcp__scrapeunblocker-remote__google_images"
    ],
    "deny": [
      "mcp__scrapeunblocker-remote__browse_page",
      "mcp__scrapeunblocker-remote__capture_xhr"
    ]
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as scrapeunblocker-remote. Review it before you use it.

Capability tags behind 25 rules
  • mcp__scrapeunblocker-remote__fetch_html private-data, untrusted-input, outbound
  • mcp__scrapeunblocker-remote__browse_page private-data, untrusted-input
  • mcp__scrapeunblocker-remote__capture_xhr untrusted-input
  • mcp__scrapeunblocker-remote__fetch_parsed private-data, untrusted-input, outbound
  • mcp__scrapeunblocker-remote__google_search private-data, untrusted-input
  • mcp__scrapeunblocker-remote__list_elements private-data, untrusted-input, outbound
  • mcp__scrapeunblocker-remote__fetch_markdown private-data, untrusted-input, outbound
  • mcp__scrapeunblocker-remote__search_and_read private-data, untrusted-input
  • mcp__scrapeunblocker-remote__list_plugins untrusted-input
  • mcp__scrapeunblocker-remote__run_plugin untrusted-input
  • mcp__scrapeunblocker-remote__amazon_search untrusted-input
  • mcp__scrapeunblocker-remote__amazon_product private-data, untrusted-input, outbound
  • mcp__scrapeunblocker-remote__amazon_reviews untrusted-input, outbound
  • mcp__scrapeunblocker-remote__walmart_search untrusted-input
  • mcp__scrapeunblocker-remote__ebay_search untrusted-input
  • mcp__scrapeunblocker-remote__google_shopping untrusted-input
  • mcp__scrapeunblocker-remote__google_local untrusted-input
  • mcp__scrapeunblocker-remote__maps_place_details untrusted-input, outbound
  • mcp__scrapeunblocker-remote__google_images untrusted-input
  • mcp__scrapeunblocker-remote__youtube_search untrusted-input
  • mcp__scrapeunblocker-remote__youtube_transcript untrusted-input
  • mcp__scrapeunblocker-remote__zillow_search untrusted-input
  • mcp__scrapeunblocker-remote__indeed_jobs untrusted-input
  • mcp__scrapeunblocker-remote__trustpilot_reviews untrusted-input
  • mcp__scrapeunblocker-remote__booking_hotels untrusted-input

Pin

Keep the decision: fail CI when the tools change, follow every change, or show a badge.

Pin in CI

Fail a build when the tools you reviewed are not the tools the server serves today.

curl -s "https://radar.anlyon.com/api/v1/servers/com.scrapeunblocker/scrapeunblocker-mcp-remote/verify?hash=1635a81824944e8787ae8ade8fe41d94c268b334a71766eb8f59363fbb3f5632" | jq -e '.changed == false'

Pinned hash 1635a8182494, from Nov 19, 2026, 14:18 UTC.

Watch

Get told when its tools change: a new tool, a new description, a new class.

https://radar.anlyon.com/servers/com.scrapeunblocker/scrapeunblocker-mcp-remote/changes.xml

An Atom feed. Works in any feed reader or Slack's RSS app.

Badge

Anlyon Radar badge for com.scrapeunblocker/scrapeunblocker-mcp-remote
[![Anlyon Radar](https://radar.anlyon.com/badge/com.scrapeunblocker/scrapeunblocker-mcp-remote.svg)](https://radar.anlyon.com/servers/com.scrapeunblocker/scrapeunblocker-mcp-remote)

Source: MCP Registry · repository. Synced Nov 19, 2026, 15:17 UTC.

scrapeunblocker-mcp-remote: questions and answers

Answered from the scan above, not from the publisher's description.

Is the scrapeunblocker-mcp-remote safe?

No scan can promise that. What its definitions show: 25 tools, 7 of which can change, send or delete something, and 7 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.

Can the scrapeunblocker-mcp-remote delete data?

Yes. 2 tools can delete, cancel or revoke something: browse_page and capture_xhr.

Can the scrapeunblocker-mcp-remote do things that cannot be undone?

No tool is classed as irreversible: none of them sends, pays, publishes or merges, going by their definitions.

What tools does the scrapeunblocker-mcp-remote have?

25 tools: 2 destructive, 5 write, 18 read. browse_page, capture_xhr, google_search, run_plugin, google_shopping, google_local, google_images, fetch_html and 17 more.

Does the scrapeunblocker-mcp-remote need sign-in or OAuth?

Its remote endpoint listed its tools to an anonymous client, without sign-in.

Does the scrapeunblocker-mcp-remote send data to outside services?

7 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): fetch_html, fetch_parsed, list_elements, fetch_markdown, amazon_product, amazon_reviews and 1 more.

Has the scrapeunblocker-mcp-remote changed its tools?

Yes, 10 times since Anlyon Radar started tracking it, most recently on Nov 19, 2026. Every change is on this page and in its Atom feed.