context safety score
A score of 36/100 indicates multiple risk signals were detected. This entity shows patterns commonly associated with malicious intent.
phishing
3 deceptive links where visible host does not match destination host
brand impersonation
The site (221ai.cn) uses DeepSeek's name, logo, and branding throughout—title 'DeepSeek网页版-免费在线使用入口', meta keywords referencing 'DeepSeek官网', and navigation labelled 'DeepSeek网页版入口'—while the footer explicitly disclaims any affiliation with DeepSeek's developer ('本网站与深度求索公司并无关联'). Users are deceived into believing this is an official DeepSeek portal. (location: page.html:8-9, page.html:42-44, page.html:2316)
brand impersonation
Multiple third-party AI tools are relabelled with DeepSeek branding. For example, 'Auto-GPT' (Significant-Gravitas) is listed as 'Auto-DeepSeek', 'PandaGPT' is listed as 'PandaDeepSeek', and the ChatGPT description contains inserted 'DeepSeek网页版' text. Official OpenAI and other brand favicons are used to make these listings appear authentic. (location: page.html:1069-1078, page.html:1359-1368, page.html:964-965)
malicious redirect
The primary 'DeepSeek网页版' call-to-action links (navbar, sidebar, hot-recommendations) do not go to deepseek.com but redirect to a third-party aggregator 'https://ai.108ai.com/web/?em3k/#/?fr=221ai.cn'. The tracking parameter '?fr=221ai.cn' and referral path indicate affiliate/monetization redirection. Users expecting the real DeepSeek are sent to an unaffiliated platform. (location: page.html:50, page.html:135, page.html:283, page.html:535, page.html:951)
social engineering
The site presents itself as the official 'DeepSeek官网入口' (official DeepSeek portal entry point), uses authoritative language ('DeepSeek满血版', 'DeepSeek官方网页版入口新体验'), and displays favicons of legitimate services (openai.com, deepseek.com, huaweicloud.com) next to listings that redirect elsewhere, creating false trust to drive traffic and affiliate conversions. (location: page.html:8-10, page.html:2278-2280, page.html:545, page.html:563, page.html:617)
malicious redirect
The Baidu share widget loads JavaScript over plain HTTP from 'http://bdimg.share.baidu.com/static/api/js/share.js', introducing a mixed-content/plaintext script load. This pattern can be exploited for script injection via MitM on the unencrypted channel. (location: page.html:2324)
brand impersonation
A listing for 'Sora官网' (Sora official site) describes the product as being released by 'DeepSeek公司' ('宇宙最强视频生成模型在2月15日的发布会上，DeepSeek公司揭晓了其最新的文生视频模型——Sor。'), falsely attributing OpenAI's Sora video model to DeepSeek. The outbound link correctly goes to openai.com/sora but the description is factually deceptive. (location: page.html:1943-1947)
social engineering
The site aggressively keyword-stuffs product descriptions and page metadata with brand names (deepseek, ChatGPT, kimi, 豆包, 嬴政ai) to capture search engine traffic from users looking for official products, then funnels them through affiliate links to monetized third-party platforms. (location: page.html:9, page.html:297-298, page.html:910-911)
malicious redirect
One 'AI生成PPT' card has an empty href on its direct-link button ('<a href="" class="content-url-togo"...>'), which is a broken/placeholder redirect. While not immediately malicious, it indicates the site may be staging content for future affiliate link insertion. (location: page.html:337)
curl https://api.brin.sh/domain/221ai.cnCommon questions teams ask before deciding whether to use this domain in agent workflows.
221ai.cn currently scores 36/100 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.