Sun.Security

Sun.Security develops a security platform focused on AI agents operating within organizational environments. The company addresses risks arising from autonomous agents that take actions and access data across workstations, cloud infrastructure, and SaaS applications, including tools such as Copilot Studio, AWS Bedrock, and coding assistants like Cursor. The platform is structured around three core functions. First, discovery: identifying agentic tools and the agents themselves across an organization's workstations, cloud environments, and SaaS applications. Second, posture management: analyzing each agent's intended purpose by examining its instructions, system prompt, connected tools, data access, and actual usage patterns. Third, runtime control: evaluating each tool call, connection, and resource request an agent makes against its established purpose before allowing execution. Sun.Security targets enterprise security teams, including CISOs and security engineering functions, that need visibility into agent activity that traditional security tools were not designed to monitor. Use cases referenced include preventing agents from taking unintended actions such as cross-tenant data exposure, unauthorized data extraction, or execution of unvetted connections to tools and MCPs (Model Context Protocols). The company positions its offering as covering the full lifecycle of AI agent security, from initial discovery and risk assessment through posture management and runtime protection to compliance-related oversight.