Show HN: Termaxa – I found my AI-agent gate had silently stopped gating Cursor (github.com)

Termaxa has launched an innovative tool designed to enhance the safety and reliability of AI coding agents by acting as a gatekeeper for shell commands they execute. Unlike traditional sandboxes that isolate potentially damaging commands, Termaxa provides a cooperative "windshield" that previews the impact of commands, automatically backs up data before execution, blocks hazardous actions, and escalates repeated destructive attempts. This approach allows users to leverage AI tools like Claude Code and Cursor while maintaining control over critical operations, thereby marrying automation with accountability. The significance of Termaxa lies in its ability to mitigate the risks associated with AI-driven code execution. By employing a user-defined policy system, it not only checks commands against predefined rules but also assesses the ramifications of those commands, including potential data loss. With features like command previews, automatic backups (using pg_dump for databases), and an audit log, Termaxa empowers developers to safely harness the power of AI coding agents without sacrificing oversight. This leap towards safer AI operation is particularly crucial as reliance on automated systems grows, ensuring that developers can confidently integrate AI into their workflows without fearing unintended consequences.