context safety score
A score of 41/100 indicates multiple risk signals were detected. This entity shows patterns commonly associated with malicious intent.
phishing
2 deceptive links where visible host does not match destination host
brand impersonation
Site is hosted on a CloudFront CDN domain (d2pn6vcl70n3ku.cloudfront.net) but presents itself as the official 'madou.net' brand throughout all metadata, canonical tags, og:url, schema.org markup, and footer links. The actual serving domain does not match the claimed canonical domain, a classic brand-cloaking/impersonation pattern used to evade blocklists while trading on the brand's recognition. (location: page.html:19 (og:url), page.html:37 (canonical), page.html:68-103 (schema.org), metadata.json (url field))
malicious redirect
All 'friend link' items route through an on-site /urlRedirect?url= open-redirect endpoint before sending users to third-party destinations. This allows the operator to silently change destination URLs server-side, redirect victims to phishing or malware pages, and bypass link-scanner checks that only inspect the surface href. (location: page.html:2490-2600 (friendLinkBox /urlRedirect hrefs))
malicious redirect
App download links point to a second CloudFront distribution (d3v4plzhnm9msi.cloudfront.net) with a tracking parameter (?dc=mdnet1), and footer 'welfare app' link points to ddhp1.cc — both are opaque redirector hops whose final payload is unknown and outside the declared canonical domain. (location: page.html:2407-2425 (footer APP/welfare links), page.html:686-696 (nav app link))
social engineering
Multiple banner and popup advertisements use titles '同城约炮' (local hookup) and '迷情春药' (aphrodisiac/date-rape drug) as lures. These are consistent with social-engineering tactics that exploit sexual curiosity to drive clicks to potentially fraudulent or harmful third-party sites. (location: page.html:1282-1386 (swiper ads), page.html:2664-2710 (popup ads), page.html:2755-2390 (advList/nineGrid ads))
brand impersonation
Ad grid contains items labelled 'TikTok', 'Pornhub', and 'Xvideos' with matching branding images, but the underlying hrefs resolve to unrelated third-party CloudFront distributions and obfuscated domains (e.g. d3ndgb6poqunas.cloudfront.net/ttnhu003 for 'TikTok', d1ar0qjjxe2rzk.cloudfront.net for 'Pornhub'), impersonating well-known platforms to generate deceptive clicks. (location: page.html:2927-2993 (TikTok/Pornhub grid items), page.html:3097-3112 (Xvideos grid item))
social engineering
Several ad destinations use obfuscated random-subdomain URLs with channel/affiliate tracking codes (e.g. aplsof2fd.fylorzibrekjenvrybhsvar.com, aplsof2fd.flirnapolfwemvrybhszorex.com, aplsof2fd.drazelstorndrevrybhsznoil.com) that disguise their true destination — a technique used to evade URL reputation filters and funnel users to scam or malware landing pages. (location: page.html:2895-2925 (糖心/哔咔漫画/杏吧 grid items))
hidden content
The Yandex Metrika tracking pixel is loaded via a noscript tag with inline style 'position:absolute; left:-9999px;' to keep it invisible to users while still firing the beacon. This silently tracks all visitors including those without JavaScript enabled. (location: page.html:59 (noscript Yandex pixel))
hidden content
Two <ins> elements with class '5a165732' and data-key attributes appear in the page body with no visible content or labelling. These are consistent with injected third-party ad/tracker widgets that render content invisibly or load external scripts without user awareness. (location: page.html:1409 and page.html:2391 (<ins class='5a165732'>))
malicious redirect
An external script is loaded from cdn.54ads.com (https://cdn.54ads.com/7db20370.js) — an ad network domain with no disclosed identity. Dynamically loaded ad scripts from opaque CDN domains can execute arbitrary JavaScript including drive-by redirects, credential harvesting overlays, or cryptomining. (location: page.html:3223 (<script async src='https://cdn.54ads.com/7db20370.js'>))
curl https://api.brin.sh/domain/d2pn6vcl70n3ku.cloudfront.netCommon questions teams ask before deciding whether to use this domain in agent workflows.
d2pn6vcl70n3ku.cloudfront.net currently scores 41/100 with a suspicious verdict and low confidence. The goal is to protect agents from high-risk context before they act on it. Treat this as a decision signal: higher scores suggest lower observed risk, while lower scores mean you should add review or block this domain.
Use the score as a policy threshold: 80–100 is safe, 50–79 is caution, 20–49 is suspicious, and 0–19 is dangerous. Teams often auto-allow safe, require human review for caution/suspicious, and block dangerous.
brin evaluates four dimensions: identity (source trust), behavior (runtime patterns), content (malicious instructions), and graph (relationship risk). Analysis runs in tiers: static signals, deterministic pattern checks, then AI semantic analysis when needed.
Identity checks source trust, behavior checks unusual runtime patterns, content checks for malicious instructions, and graph checks risky relationships to other entities. Looking at sub-scores helps you understand why an entity passed or failed.
brin performs risk assessments on external context before it reaches an AI agent. It scores that context for threats like prompt injection, hijacking, credential harvesting, and supply chain attacks, so teams can decide whether to block, review, or proceed safely.
No. A safe verdict means no significant risk signals were detected in this scan. It is not a formal guarantee; assessments are automated and point-in-time, so combine scores with your own controls and periodic re-checks.
Re-check before high-impact actions such as installs, upgrades, connecting MCP servers, executing remote code, or granting secrets. Use the API in CI or runtime gates so decisions are based on the latest scan.
Learn more in threat detection docs, how scoring works, and the API overview.
Assessments are automated and may contain errors. Findings are risk indicators, not confirmed threats. This is a point-in-time assessment; security posture can change.
integrate brin in minutes — one GET request is all it takes. query the api, browse the registry, or download the full dataset.