Autonomous Mission-Critical Systems & AST Safety Auditor, powered by Strands Agents SDK and Model Context Protocol (MCP).
Demonstrating: (1) Sub-15ms Power of 10 AST analysis, (2) Deterministic & entropy secret scanning, (3) Standalone audit script validation (0 violations), (4) Strands InterventionHandler confirmation gating, and (5) 85/85 tests green.
Demonstrating the hackathon mandate: the agent runs silently in the background and only interrupts the human when a true critical invariant violation or secret leak requires an executive decision.