Justin Kwon

Rust. Security tooling and AI/LLM security (AI Security Posture Management). Host-side monitoring for AI coding agents.

How AI coding agents actually get attacked: the config files (hooks, MCP allowlists, permissions) that quietly grant permission. Also sharpening my Rust.

Sigil: an open-source, host-side AI-SPM agent that watches AI coding agent config files and scores them when they turn dangerous. github.com/Ju571nK/sigil

Talking AI-SPM, agentic-AI security, prompt injection, and config-as-attack-surface. Feedback on Sigil (my OSS tool) is very welcome.