context safety score
A score of 38/100 indicates multiple risk signals were detected. This entity shows patterns commonly associated with malicious intent.
phishing
83 deceptive links where visible host does not match destination host
social engineering
The site uses a systematic bait-and-switch link deception pattern across 83+ links: each anchor's href, title attribute, and visible link text all point to three completely different, unrelated topics. This is designed to manipulate search engine crawlers and deceive users/agents about the actual content of destination pages. This is a large-scale SEO link farm operation. (location: page.html: lines 68-77, 107-118, 179-274, and throughout entire page body)
brand impersonation
The site at downg.com presents itself as '绿软家园' (a well-known Chinese green/free software download brand) in its title, meta tags, logo, and copyright notice, but serves entirely different English-learning blog content. All backend infrastructure (WordPress API, RSS feed, xmlrpc.php, wlwmanifest.xml, comment endpoints) points to www.jiandalou.net, indicating the site is impersonating or piggybacking on an established brand identity while operating under a different domain. (location: page.html: lines 9-11, 20-21, 48, 59, 1720-1722; metadata.json)
malicious redirect
The search form action is set to 'http://www.jiandalou.net' (off-domain, HTTP not HTTPS), meaning any user search query submitted on downg.com is sent to a different third-party domain. This constitutes an off-domain form action that can harvest search queries and redirect users away from the displayed site. (location: page.html: line 84)
social engineering
Article timestamps are fabricated: dozens of articles are dated '2021年11月' or '2021年12月' but contain content clearly referencing 2025 events (e.g., '2025上海春节消费券', SpaceX Starship S35 2025 explosion, '2025年春节联欢晚会'). This timestamp spoofing is used to manipulate content freshness signals and deceive crawlers/agents about content age and authenticity. (location: page.html: lines 117, 133, 149, 189, 205, 221, 237, 253, 288, 304, 320, 336, 352, 368 and throughout)
hidden content
Navigation menu items use deceptive title attributes that describe completely different content from the visible link text. For example, the '零基础英语' nav link has a title describing a government environmental enforcement meeting. This hides real link destinations from users while potentially feeding different content to AI agents or scrapers that parse title attributes. (location: page.html: lines 68-76)
social engineering
A malformed href attribute contains an extra single-quote character: href="'https://downg.com/11536926.html'", indicating automated/programmatic page generation with poor quality control, consistent with a large-scale spam/SEO farm operation rather than a legitimate site. (location: page.html: line 1552)
curl https://api.brin.sh/domain/downg.comCommon questions teams ask before deciding whether to use this domain in agent workflows.
downg.com currently scores 38/100 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.