context safety score
A score of 44/100 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
phishing
1 deceptive links where visible host does not match destination host
malicious redirect
Multiple internal content links (thematique slider, video section, events section) use hardcoded private IP address http://172.19.34.167/ instead of the public domain. This leaks internal server IP and could redirect users to an internal/attacker-controlled host if the IP is reachable from the client network. Approximately 14 anchor hrefs point to http://172.19.34.167/... (location: page.html lines 1249-1460, 1654, 1688, 1717 — thematique-item anchors, video-item anchors, event-item anchors)
malicious redirect
External link to http://prediagentreprise.fr/ (plain HTTP, no HTTPS) embedded as a trusted tool link inside the Bpifrance Création site. The domain is a separate third-party site loaded over unencrypted HTTP, enabling MITM attacks or substitution of a malicious page presented under the Bpifrance brand. (location: page.html lines 686 and 1038 — outil-item anchor href="http://prediagentreprise.fr/")
hidden content
A chatbot is dynamically injected at runtime via JavaScript by constructing a <script> element pointing to https://chatbot.synapse-developpement.fr/chatbotuiLauncher/config.aspx. The bot credentials (botKey: '05j8daHhuX3nroyPWTlKsbKV78q6vcq6GPMHF2Oe', synapseBotId: 'Bpifrance-creation-llm-V2') are exposed in plaintext in the page source and the script is loaded from a third-party domain not under bpifrance-creation.fr control. This creates a supply-chain vector for injecting arbitrary content into the page. (location: page.html lines 2070-2079 and page-text.txt lines 1970-1978 — window.botSettings / dynamic script injection block)
prompt injection
The exposed chatbot integration (synapseBotId: 'Bpifrance-creation-llm-V2') loaded from chatbot.synapse-developpement.fr introduces an LLM-backed assistant into the page. A compromised or malicious third-party chatbot script could inject adversarial prompts into conversations handled by AI agents browsing or summarising this page, manipulating agent behaviour under the guise of the official Bpifrance assistant. (location: page.html lines 2070-2079 — window.botSettings block with third-party LLM chatbot script src)
hidden content
The body element carries class='entrepreneur d-none' on initial load, hiding the entire page body until JavaScript runs. Combined with the JS-injected chatbot and dynamic content, meaningful page content is invisible without JavaScript execution, which can be used to serve different content to crawlers/agents vs. human browsers. (location: page.html line 102 — <body class="entrepreneur d-none" id="bpi-espace-entrepreneur">)
curl https://api.brin.sh/domain/bpifrance-creation.frCommon questions teams ask before deciding whether to use this domain in agent workflows.
bpifrance-creation.fr currently scores 44/100 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.