Sentinel — Credibility Engine

Sentinel doesn’t wait for malware — it detects the deception that leads you there. Sentinel is a deception detection engine designed to identify scam patterns, brand impersonation, and manipulative narratives before you interact with a site. Unlike traditional security tools that respond after a threat is triggered, Sentinel works earlier in the chain — analysing structure, behaviour, and intent to flag risk before credentials are entered, downloads begin, or decisions are made. This means avoiding the threat entirely, rather than reacting after exposure. The engine evaluates websites using three layers of analysis: structural signals from the domain, narrative patterns found in page content, and behavioural conflict patterns commonly used in fraudulent campaigns. Sentinel runs silently in the background and performs all analysis locally inside your browser. No browsing data is collected, transmitted, or stored externally. This project is currently in active development and is being released free to use while the detection engine continues to improve. Users are encouraged to submit feedback and help refine detection accuracy and reduce false positives.