context safety score
A score of 24/100 indicates multiple risk signals were detected. This entity shows patterns commonly associated with malicious intent.
encoded payload
suspicious base64-like blobs detected in page content
brand impersonation
The page at flushpersist.com (a non-Google domain, 246 days old) fully impersonates a Google CAPTCHA/security page. The HTML title is set to 'https://google.com/', the body text references 'Google automatically detects requests' and 'Terms of Service', and a real Google reCAPTCHA Enterprise widget is embedded to lend authenticity. The page is designed to appear as though it originates from Google. (location: page.html:3, page.html:24, page.html:27)
phishing
The form (id='captcha-form') posts to 'index' on flushpersist.com with a hidden 'continue' value of 'https://google.com/' and an opaque token in hidden field 'q'. Victims who complete the fake CAPTCHA submit their interaction/token data to the attacker-controlled domain, not Google. This is a classic phishing interstitial that harvests CAPTCHA responses and redirects victims. (location: page.html:7, page.html:17)
malicious redirect
A hidden input field named 'continue' carries the value 'https://google.com/', and the form posts to 'index' on the attacker domain. After CAPTCHA submission, the victim is likely redirected to google.com to conceal the attack, while the attacker silently logs the interaction or token. This redirect is designed to mask the phishing flow. (location: page.html:17)
social engineering
The page employs authoritative Google language ('Our systems have detected unusual traffic', 'Terms of Service', 'solving the above CAPTCHA will let you continue to use our services') to pressure users into completing a CAPTCHA on a third-party attacker-controlled domain. The collapsible 'Why did this happen?' section adds false legitimacy and urgency. (location: page.html:24-28, page-text.txt:21-24)
prompt injection
The page title is set to 'https://google.com/' and all visible text references Google infrastructure and policies. An AI agent browsing or summarizing this page would likely report it as a legitimate Google page, causing it to misclassify the domain and potentially follow or trust the embedded redirect instructions. This constitutes indirect prompt injection targeting AI agents that read page content. (location: page.html:3, page.html:24, page.html:27, page.html:30)
hidden content
A hidden input field 'q' contains a long opaque base64-like token value ('EhAmABkAAAAtBQAAAAAAADgBGLPO...') not visible to users. Its purpose is undisclosed and it is submitted silently with the form, potentially encoding tracking identifiers, session data, or anti-analysis markers. (location: page.html:17)
credential harvesting
The reCAPTCHA Enterprise sitekey ('6LfwuyUTAAAAAOAmoS0fdqijC2PbbdH4kjq62Y1b') combined with the data-s parameter and the hidden 'q' token suggests infrastructure for harvesting verified-human CAPTCHA tokens. These tokens can be resold or replayed to bypass Google CAPTCHA protections on other platforms, constituting indirect credential/token harvesting. (location: page.html:15, page.html:17)
curl https://api.brin.sh/domain/flushpersist.comCommon questions teams ask before deciding whether to use this domain in agent workflows.
flushpersist.com currently scores 24/100 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.