AgentCops – SCAB AI Risk Scanner

Scan pages and AI conversations using the SCAB protocol. Save scan history locally and unlock premium features via agentcops.com.

AgentCops is a browser-based AI risk monitoring extension designed to help users, organizations, and professionals assess the behavioral safety of AI-generated content and high-risk webpages in real time. AgentCops applies the SCAB Protocol — a structured behavioral assessment framework — to scan visible page content and AI interactions, generating clear, interpretable risk signals without collecting personal data or transmitting browsing history. This extension is built for transparency, safety, and human oversight. ━━━━━━━━━━━━━━━━━━━━━━ WHAT IS SCAB? ━━━━━━━━━━━━━━━━━━━━━━ SCAB is a six-domain behavioral evaluation framework used to assess potential AI risk signals: • Sovereignty — Does the content attempt to override user autonomy or pressure decisions? • Coherence — Does the content show instability, contradiction, or breakdown in reasoning? • Alignment — Does the content promote unethical, unsafe, or misaligned actions? • Boundaries — Does it cross safety, security, or data-protection boundaries? • Agency — Does it reduce user choice, encourage dependency, or manipulate behavior? • Probity — Does it show signs of deception, misinformation, or unverifiable claims? Each domain is scored from 0–3, producing a total risk score from 0–18. Lower scores indicate safer behavior. Higher scores indicate elevated or critical risk signals. ━━━━━━━━━━━━━━━━━━━━━━ CORE FEATURES ━━━━━━━━━━━━━━━━━━━━━━ 🔍 One-Click AI & Page Scanning Scan the currently visible webpage or AI conversation instantly from your browser toolbar. 📊 SCAB Risk Breakdown View individual scores for all six SCAB domains plus a total risk score and severity level: • Low • Moderate • High • Critical 🧠 Human-Readable Results AgentCops does not “black box” results. Scores are presented clearly so users can apply judgment and context. 🗂 Local Scan History (Optional) Store recent scans locally on your device for review and comparison. History never leaves your browser unless you choose to export it elsewhere. 🚨 Risk Alert Badge Optionally display an alert badge when scan results exceed your chosen risk threshold. ⚙️ Configurable Thresholds Adjust alert sensitivity and history retention to match your workflow or organizational policy. ━━━━━━━━━━━━━━━━━━━━━━ PRIVACY & DATA HANDLING ━━━━━━━━━━━━━━━━━━━━━━ AgentCops is privacy-first by design. • No browsing history is uploaded • No page content is transmitted by default • No personal data is collected • No advertising or tracking • All scans run locally inside the browser Premium licensing verification only checks an anonymous license token and does not include browsing data or scan content. ━━━━━━━━━━━━━━━━━━━━━━ PREMIUM FEATURES (OPTIONAL) ━━━━━━━━━━━━━━━━━━━━━━ AgentCops offers optional premium functionality unlocked via purchase on agentcops.com: • Advanced monitoring packs • Extended configuration controls • Priority feature access as the platform evolves Premium purchases are handled securely on our website using Stripe. The extension itself never stores payment information. ━━━━━━━━━━━━━━━━━━━━━━ WHO THIS IS FOR ━━━━━━━━━━━━━━━━━━━━━━ AgentCops is designed for: • Professionals working with AI tools • Developers and engineers • Researchers and analysts • Compliance and governance teams • Educators and students • Journalists and investigators • Anyone concerned about AI behavior, manipulation, or misuse ━━━━━━━━━━━━━━━━━━━━━━ IMPORTANT NOTES ━━━━━━━━━━━━━━━━━━━━━━ • AgentCops is an assistive risk-awareness tool, not a legal or compliance authority. • Results should be interpreted alongside human judgment and contextual knowledge. • The extension does not replace professional review, audits, or formal governance processes. ━━━━━━━━━━━━━━━━━━━━━━ TRANSPARENCY & FUTURE DEVELOPMENT ━━━━━━━━━━━━━━━━━━━━━━ AgentCops is part of a broader initiative focused on responsible AI deployment, behavioral governance, and human-centered oversight. Future updates may include: • Expanded SCAB analysis depth • Enterprise dashboards • Audit export tools • Team and organizational views • Policy-aligned reporting modes ━━━━━━━━━━━━━━━━━━━━━━ WEBSITE ━━━━━━━━━━━━━━━━━━━━━━ https://agentcops.com