Writing

Notes on AI agent security, LangGraph hardening, and the parts of building with LLMs that don't get enough attention.