context safety score
A score of 32/100 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 hosted on sourshaped.com renders a full replica of a Google CAPTCHA/interstitial page. The <title> is set to 'https://google.com/', the body text impersonates Google's automated-traffic detection system, references Google's Terms of Service, and uses Google's support URLs. The actual domain is sourshaped.com, not google.com. (location: page.html:3, page.html:22-28)
phishing
The page mimics a legitimate Google security checkpoint to deceive users into believing they are interacting with Google infrastructure. A hidden form posts to 'index' with a 'continue' value of 'https://google.com/', creating a convincing phishing flow that captures CAPTCHA interaction and then redirects to Google to avoid suspicion. (location: page.html:7, page.html:17)
malicious redirect
A hidden input field named 'continue' is set to 'https://google.com/' and submitted via the CAPTCHA form. This is a classic redirect-after-capture pattern: once the victim solves the CAPTCHA, the form posts to the attacker-controlled 'index' endpoint (on sourshaped.com) before redirecting to Google, allowing the attacker to log the interaction server-side. (location: page.html:17)
credential harvesting
The form contains an opaque hidden field 'q' with a long base64-like encoded value, submitted via POST to 'index' on sourshaped.com. This encoded token likely identifies the victim session and is harvested server-side on form submission, potentially used to track or de-anonymize the user. (location: page.html:17)
social engineering
The page uses authoritative Google-branded language ('Our systems have detected unusual traffic', 'violation of the Terms of Service', 'solving the above CAPTCHA will let you continue') to pressure the user into completing the CAPTCHA form, a classic urgency-and-authority social engineering technique deployed on a non-Google domain. (location: page.html:24-28, page-text.txt:21-24)
prompt injection
The page title is set to the string 'https://google.com/' rather than a descriptive title. AI agents that index or summarize page titles may interpret this as a legitimate Google URL, causing the agent to misclassify the page origin or follow the embedded URL as a trusted destination. This is a prompt injection vector targeting URL-aware AI agents. (location: page.html:3)
hidden content
An 'infoDiv' element is rendered with style='display:none' by default and only revealed on user click. This div contains additional Google-branded text and links designed to reinforce the impersonation. Hidden content is used to pass simple visibility checks while still influencing users who interact with the page. (location: page.html:26-28)
curl https://api.brin.sh/domain/sourshaped.comCommon questions teams ask before deciding whether to use this domain in agent workflows.
sourshaped.com currently scores 32/100 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.