AgentSure PTE.LTD. Information

AgentSure builds JANUS, a platform for securing AI agents, starting with the business rules they are supposed to follow. Most guardrails block generic harms. They don't know your refund limits, approval thresholds or data-access rules, because those rules live in prompts and code, and are rarely written down or tested. JANUS closes that gap: • JANUS Audit: recovers the business rules (the SPEC) your agent should follow from its code, prompts and tools, tests every one, and maps your AI supply chain (AI-BOM). • JANUS Pentest: simulated attacks on agents and their AI components, including prompt injection, tool hijacking and MCP/A2A abuse. Evidence, not a score. • JANUS Runtime Monitor (Early Access): enforces the SPEC and guard rules on every live action: allow, block, escalate, log. Available as SaaS or on-premises. Where we come from AgentSure is a CRPO / SMU spin-off. The research behind JANUS was funded by the CyberSG R&D Programme Office (CRPO) under the Cyber Security Agency of Singapore (CSA). JANUS Audit and JANUS Pentest are deployed at a national government agency. AgentSure is a General Member of the AI Verify Foundation. Co-founded by Tiger Li, an actuary (FSA, FCAA, CERA), and Asst Prof Xiaofei Xie of Singapore Management University, a ten-time recipient of the ACM SIGSOFT Distinguished Paper Award. Measured, managed risk is what makes AI insurable. The same JANUS engine is built to support insurers in underwriting AI agent risk. support@agentsure.tech AgentSure Pte. Ltd. · Singapore