Harden AIF (Agentic Integrity Foundation) is a local cybersecurity tool that intercepts and judges every tool call made by AI coding agents before execution. It runs proprietary cybersecurity-focused language models on-device to allow, block, redact, or safely redirect agent actions. It supports popular coding agents including Claude Code, Codex, Cursor, and Kiro through native hooks or an MCP proxy fallback. The free tier requires no account and includes core pre-execution blocking, a local audit log, and telemetry opt-out, with enterprise options available for compliance reporting and managed deployment.