WebLens MCP server
dev.weblens/weblensweblens.devstreamable-httpv2.1.0Scanned
Scrape, crawl, map and extract the web. Pay per call in USDC, no account or API key.
26 tools: 1 do something that can't be undone, 12 change things and 13 only read.
- Reads private data · 1
- Reads text others write · 8
- Can send data out · 10
- 1 finding
Tools listed without sign-in. Remote posture
With Anlyon
1 tool here does something that can't be undone. Govern it.
- Anlyon lists the tools and adopts none.
- Each tool you adopt gets approvals and receipts.
- Your agent never holds the key.
A tool your agent calls directly is not routed through Anlyon and is not governed by it. Anlyon Radar only lists and labels, and these labels are the same either way.
What the WebLens MCP server can do
The WebLens MCP server is published by weblens.dev, a domain the MCP Registry verified. It has 26 tools. It can do things that cannot be undone, such as sending, paying or publishing (screenshot) and create or change data (12 tools: map_site, crawl_site, extract_pdf, extract_data and 8 more), but cannot delete data. 13 tools only read.
1 tool reads private data (search_discussions); 8 tools bring in text other people can write (preview_endpoint, fetch_webpage, fetch_webpage_pro and 5 more); 10 tools can send data out of your system (preview_endpoint, fetch_webpage, fetch_webpage_pro and 7 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 1 finding (base64-blob). Each one comes with the text it matched.
Its remote endpoint lists its tools without sign-in, and its tools changed once since Anlyon Radar first saw them.
Last scanned Nov 11, 2026, version 2.1.0.
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.
26 tools
Can't be undone 1Claude Code: ask
screenshotCapture high-quality screenshots of any webpage using headless browser.Reads text others writeCan send data outNo findings
Why it is in this group
- The name "screenshot" says "screenshot": a word for reading.
- The description's first verb is "capture": a word that usually means money moves, which cannot be taken back.
What it can reach
- untrusted-inputThe description's first sentence says "capture" and "webpage": 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.
Description, as the server wrote it Capture high-quality screenshots of any webpage using headless browser. Price: $0.008
Changes things 12Claude Code: ask
answer_questionGrounded answer with inline [n] citations: searches the web, fetches sources, and answers strictly from them.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it Grounded answer with inline [n] citations: searches the web, fetches sources, and answers strictly from them. Price: $0.05
company_profileA company profile from its domain, read live from its own site: what it sells and to whom, pricing and plans, NAICS sector, self-reported head count band, open roles from its Green…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it A company profile from its domain, read live from its own site: what it sells and to whom, pricing and plans, NAICS sector, self-reported head count band, open roles from its Greenhouse/Lever/Ashby board, tech stack, socials, security.txt role contacts, and the domain's registrar, age, email provider, SaaS vendors and SPF/DMARC posture. Each AI-read field names its source page; coverage flags show what was found. Price: $0.05 (70% off when served from cache)
crawl_siteCrawl a site and get clean markdown for every page, in one call.Reads text others writeCan send data outNo findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
What it can reach
- untrusted-inputThe name "crawl_site" says "crawl": it reads web pages, which anyone can write.
- untrusted-inputThe description's first sentence says "crawl": 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.
Description, as the server wrote it Crawl a site and get clean markdown for every page, in one call. Same-host BFS with depth and page-budget limits, robots.txt honoured by default. $0.0015 per requested page (1-25). Price: $0.0015 per requested page (1-25); an x402 payment is at least $0.004
deep_researchMulti-step research in one call: plans sub-questions, searches each, fetches and dedupes sources, and synthesizes an answer with inline [n] citations, key findings, and gaps.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it Multi-step research in one call: plans sub-questions, searches each, fetches and dedupes sources, and synthesizes an answer with inline [n] citations, key findings, and gaps. Tiers: quick = 1 search / 5 sources ($0.08); standard = 3 sub-questions / 8 sources ($0.20); deep = 5 sub-questions / 12 sources ($0.35). Default: standard. Price: quick = 1 search / 5 sources ($0.08); standard = 3 sub-questions / 8 sources ($0.20); deep = 5 sub-questions / 12 sources ($0.35). Default: standard.
detect_techWhat a site is built and run on, from a single fetch: frameworks, CMS, ecommerce platform, CDN, analytics, payments, support widgets and web server — each with the header or HTML m…Can send data outNo findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
What it can reach
- outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
Description, as the server wrote it What a site is built and run on, from a single fetch: frameworks, CMS, ecommerce platform, CDN, analytics, payments, support widgets and web server — each with the header or HTML marker that proves it. Price: $0.005
domain_intelDomain intelligence in one call: RDAP registration (registrar, age, expiry, status), live DNS (A/AAAA/MX/NS/TXT), the mail and DNS providers behind them, the SaaS vendors the domai…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it Domain intelligence in one call: RDAP registration (registrar, age, expiry, status), live DNS (A/AAAA/MX/NS/TXT), the mail and DNS providers behind them, the SaaS vendors the domain's TXT verification tokens reveal, SPF/DMARC posture, and risk signals like newly-registered or no-registrar-lock. Price: $0.005
extract_dataExtract structured data from any webpage using JSON schema.Can send data outNo findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
What it can reach
- outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
Description, as the server wrote it Extract structured data from any webpage using JSON schema. AI-powered extraction that understands page context. Price: $0.03, up to 3x on bot-protected or complex sites
extract_pdfExtract text and metadata from PDF documents.Can send data outNo findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
What it can reach
- outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
Description, as the server wrote it Extract text and metadata from PDF documents. Supports page-specific extraction and returns structured content. Price: $0.004
map_siteDiscover a site's URLs without fetching page content.Can send data outNo findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
What it can reach
- outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
Description, as the server wrote it Discover a site's URLs without fetching page content. Reads robots.txt sitemap directives, sitemap.xml and nested sitemap indexes, falling back to homepage link extraction. Price: $0.004
package_intelShould you depend on this package?No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it Should you depend on this package? Version, license, deprecation (with the maintainer's reason), weekly and monthly downloads, last release date, maintainer count, npm quality/popularity/maintenance scores, and health signals — npm or PyPI, one call. Price: $0.003; an x402 payment is at least $0.004
project_due_diligenceOff-chain due diligence on a project's web presence — the half on-chain rug checkers cannot see.1 finding
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Findings
- Warn
The input schema at properties.tokenAddress.example contains a 42-character base64 string.
What this check looks forMatched text
inputSchema.properties.tokenAddress.example: "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913"
Description, as the server wrote it Off-chain due diligence on a project's web presence — the half on-chain rug checkers cannot see. Domain age and registrar, whether a team page, whitepaper, docs and socials exist, whether the site is an off-the-shelf template, and a cross-check of the contract address you are about to trade against the addresses actually printed on the project's own website. Returns weighted risk signals and an A-F grade. Reads nothing on-chain. Price: $0.05
youtube_transcriptFull transcript of any YouTube video with timestamps.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
Description, as the server wrote it Full transcript of any YouTube video with timestamps. Accepts a video ID or any YouTube URL. Price: $0.03
Only reads 13Show 13 read-only tools
fetch_webpageFetch and convert any webpage to clean markdown.Reads text others writeCan send data outNo findings
Why it is in this group
- The name "fetch_webpage" says "fetch": a word for reading.
What it can reach
- untrusted-inputThe name "fetch_webpage" says "fetch" and "webpage": it reads web pages, which anyone can write.
- untrusted-inputThe description's first sentence says "fetch" and "webpage": 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 and convert any webpage to clean markdown. Fast, no JavaScript rendering. Perfect for static content, articles, and documentation. Price: $0.002 (70% off when served from cache); an x402 payment is at least $0.004
fetch_webpage_proFetch webpage with full JavaScript rendering using headless browser.Reads text others writeCan send data outNo findings
Why it is in this group
- The name "fetch_webpage_pro" says "fetch": a word for reading.
What it can reach
- untrusted-inputThe name "fetch_webpage_pro" says "fetch" and "webpage": it reads web pages, which anyone can write.
- untrusted-inputThe description's first sentence says "fetch" and "webpage": 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 webpage with full JavaScript rendering using headless browser. Perfect for SPAs, React/Vue apps, and dynamic content that requires JS execution. Price: $0.006, up to 3x on bot-protected or complex sites (70% off when served from cache); an x402 payment is at least $0.004
get_contentsCheap bulk page text: fetch 1-20 URLs and get clean markdown, truncated to a per-page cap.Reads text others writeCan send data outNo findings
Why it is in this group
- The name "get_contents" says "get": a word for reading.
What it can reach
- untrusted-inputThe input takes a URL (input.urls), and the tool reads it: it reads web pages, which anyone can write.
- outboundThe input takes a URL (input.urls): 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 Cheap bulk page text: fetch 1-20 URLs and get clean markdown, truncated to a per-page cap. $0.0015 per URL. Price: $0.0015 per URL (1-20); an x402 payment is at least $0.004
preview_endpointFREE.Reads text others writeCan send data outNo findings
Why it is in this group
- The name "preview_endpoint" says "preview": a word for reading.
What it can reach
- untrusted-inputThe input takes a URL (input.endpoint), and the tool reads it: it reads web pages, which anyone can write.
- outboundThe input takes a URL (input.endpoint): 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 FREE. See what a paid WebLens endpoint costs and what it returns before paying: the live price, a one-line summary, and a recorded sample of the exact response shape. Endpoints with no paid upstream (/fetch/basic, /contents, /map) also run a real truncated LIVE preview when you pass a url; search- and Anthropic-backed endpoints return the recorded sample only, because free live runs there would burn upstream credits. Price: free
search_autocompleteGoogle Autocomplete suggestions for a partial query.No findings
Why it is in this group
- The name "search_autocomplete" says "search": a word for reading.
Description, as the server wrote it Google Autocomplete suggestions for a partial query. Useful for keyword research and intent discovery. Price: $0.004
search_discussionsWhat Hacker News said about a topic: matching stories with points, comment counts and links to the threads, plus aggregates — total matches, points and comments returned, the domai…Reads private dataReads text others writeNo findings
Why it is in this group
- The name "search_discussions" says "search": a word for reading.
What it can reach
- private-dataThe description's first sentence says "thread": it reads messages.
- untrusted-inputThe name "search_discussions" says "search" and "discussion": it reads comments, which other people can write.
- untrusted-inputThe description's first sentence says "comment": it reads comments, which other people can write.
Description, as the server wrote it What Hacker News said about a topic: matching stories with points, comment counts and links to the threads, plus aggregates — total matches, points and comments returned, the domains most often submitted, and the first/last time it was discussed. Price: $0.004
search_imagesGoogle Images search.No findings
Why it is in this group
- The name "search_images" says "search": a word for reading.
Description, as the server wrote it Google Images search. Returns direct image URLs with dimensions, thumbnails, and source pages. Price: $0.005
search_newsReal-time news search via Google News.No findings
Why it is in this group
- The name "search_news" says "search": a word for reading.
Description, as the server wrote it Real-time news search via Google News. Returns up to 10 ranked articles with snippet, source, date, and thumbnail. Price: $0.005
search_placesLocal business search via Google Places.No findings
Why it is in this group
- The name "search_places" says "search": a word for reading.
Description, as the server wrote it Local business search via Google Places. Returns names, addresses, categories, ratings, reviews, phone numbers, websites, and coordinates. Price: $0.045
search_scholarGoogle Scholar academic search.No findings
Why it is in this group
- The name "search_scholar" says "search": a word for reading.
Description, as the server wrote it Google Scholar academic search. Returns papers with snippets, publication info, and citation counts. Price: $0.005
search_shoppingGoogle Shopping product search.No findings
Why it is in this group
- The name "search_shopping" says "search": a word for reading.
Description, as the server wrote it Google Shopping product search. Returns products with prices, sellers, ratings, and links. Price: $0.005
search_trendsGoogle Trends interest over time for a topic, worldwide: a timeline of relative search interest per period (0-100, where 100 is the topic's peak).No findings
Why it is in this group
- The name "search_trends" says "search": a word for reading.
Description, as the server wrote it Google Trends interest over time for a topic, worldwide: a timeline of relative search interest per period (0-100, where 100 is the topic's peak). Use it to see whether interest in a product, technology, brand or keyword is rising, falling or seasonal. Price: $0.04
search_webReal-time web search powered by Google.Reads text others writeNo findings
Why it is in this group
- The name "search_web" says "search": a word for reading.
What it can reach
- untrusted-inputThe name "search_web" says "search" and "web": it reads web pages, which anyone can write.
- untrusted-inputThe description's first sentence says "web": it reads web pages, which anyone can write.
Description, as the server wrote it Real-time web search powered by Google. Returns ranked results with titles, URLs, and snippets. Set includeContent to also fetch the top result pages as markdown in the same call. Price: $0.005, plus $0.0015 per result page fetched with includeContent
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.
80 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 1 private-data tool breaks every one.
- Reads private data1
- Reads files, mail, messages, databases, secrets, calendars or repositories.
- Reads text others write8
- Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
- Can send data out10
- Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
{
"permissions": {
"deny": [
"mcp__weblens__search_discussions"
]
}
}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.
Looked at Nov 20, 2026, 18:53 UTC, unchanged since Oct 22, 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
- 13
- Protocol version
- 2025-06-18
- WWW-Authenticate on 401
- not observed
- resource_metadata in WWW-Authenticate
- not observed
- RFC 9728 protected resource metadata
- not observed
- Authorization servers
- not observed
- Legacy HTTP+SSE transport
- no
- Dynamic client registration
- not observed
- Client ID metadata documents
- not observed
- HighThe whole server
The endpoint served tools/list without credentials, and the list has 13 tools that change things (1 irreversible, 12 write). Whether a call also works without credentials was not tested: tools are never called.
What this check looks forMatched text
https://api.weblens.dev/mcp: tools/list returned 26 tools to an anonymous client, including screenshot (irreversible), map_site (write), crawl_site (write), extract_pdf (write), extract_data (write) and 8 more
Findings
Matches of fixed checks, with the text they matched.
The input schema at properties.tokenAddress.example contains a 42-character base64 string.
What this check looks forMatched text
inputSchema.properties.tokenAddress.example: "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913"
History
Changes to its tool definitions and to its remote endpoint, newest first.
Tools1 removed, 9 changed.−1 removed~9 changedView diff
- Removed:fetch_resilient
- Changed:company_profileChanges thingsdescription
- Changed:preview_endpointOnly readsdescription
- Changed:search_autocompleteOnly readsdescription
- Changed:search_imagesOnly readsdescription
- Changed:search_newsOnly readsdescription
- Changed:search_placesOnly readsdescription
- Changed:search_scholarOnly readsdescription
- Changed:search_shoppingOnly readsdescription
- Changed:search_webOnly readsdescription
Install
Pinned to version 2.1.0. Values in angle brackets are yours to fill in.
claude mcp add --transport http weblens https://api.weblens.dev/mcpclaude mcp add --transport http weblens https://api.weblens.dev/mcpRun 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.
{
"permissions": {
"allow": [
"mcp__weblens__search_web",
"mcp__weblens__search_news",
"mcp__weblens__search_images",
"mcp__weblens__search_places",
"mcp__weblens__search_shopping",
"mcp__weblens__search_scholar",
"mcp__weblens__search_autocomplete",
"mcp__weblens__search_trends",
"mcp__weblens__search_discussions"
],
"ask": [
"mcp__weblens__preview_endpoint",
"mcp__weblens__fetch_webpage",
"mcp__weblens__fetch_webpage_pro",
"mcp__weblens__screenshot",
"mcp__weblens__get_contents",
"mcp__weblens__map_site",
"mcp__weblens__crawl_site",
"mcp__weblens__extract_pdf",
"mcp__weblens__extract_data",
"mcp__weblens__youtube_transcript",
"mcp__weblens__answer_question",
"mcp__weblens__deep_research",
"mcp__weblens__domain_intel",
"mcp__weblens__package_intel",
"mcp__weblens__detect_tech",
"mcp__weblens__company_profile",
"mcp__weblens__project_due_diligence"
],
"deny": []
}
}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 weblens. Review it before you use it.
Capability tags behind 12 rules
- mcp__weblens__preview_endpoint untrusted-input, outbound
- mcp__weblens__fetch_webpage untrusted-input, outbound
- mcp__weblens__fetch_webpage_pro untrusted-input, outbound
- mcp__weblens__screenshot untrusted-input, outbound
- mcp__weblens__get_contents untrusted-input, outbound
- mcp__weblens__map_site outbound
- mcp__weblens__crawl_site untrusted-input, outbound
- mcp__weblens__extract_pdf outbound
- mcp__weblens__extract_data outbound
- mcp__weblens__search_web untrusted-input
- mcp__weblens__search_discussions private-data, untrusted-input
- mcp__weblens__detect_tech outbound
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/dev.weblens/weblens/verify?hash=7d8382bea9e59e9bfc5958e08f6cda335d0daa03d780050b7b44c006f2c534e5" | jq -e '.changed == false'Pinned hash 7d8382bea9e5, from Nov 11, 2026, 10:18 UTC.
Watch
Get told when its tools change: a new tool, a new description, a new class.
https://radar.anlyon.com/servers/dev.weblens/weblens/changes.xmlAn Atom feed. Works in any feed reader or Slack's RSS app.
Badge
[](https://radar.anlyon.com/servers/dev.weblens/weblens)Source: MCP Registry · repository. Synced Oct 21, 2026, 18:12 UTC.
WebLens MCP server: questions and answers
Answered from the scan above, not from the publisher's description.
Is the WebLens MCP server safe?
- No scan can promise that. What its definitions show: 26 tools, 13 of which can change, send or delete something, and 1 finding from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.
Can the WebLens MCP server delete data?
- No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.
Can the WebLens MCP server do things that cannot be undone?
- Yes. 1 tool can send, pay, publish or merge: screenshot.
What tools does the WebLens MCP server have?
- 26 tools: 1 irreversible, 12 write, 13 read. screenshot, map_site, crawl_site, extract_pdf, extract_data, youtube_transcript, answer_question, deep_research and 18 more.
Does the WebLens MCP server need sign-in or OAuth?
- Its remote endpoint listed its tools to an anonymous client, without sign-in.
Does the WebLens MCP server send data to outside services?
- 10 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): preview_endpoint, fetch_webpage, fetch_webpage_pro, screenshot, get_contents, map_site and 4 more.
Has the WebLens MCP server changed its tools?
- Yes, once since Anlyon Radar started tracking it, most recently on Oct 29, 2026. Every change is on this page and in its Atom feed.