Posts tagged: ai-safety

How a Single ".env" in allowed-extensions Could Leak Your Secrets to LLMs

A three-layer defense-in-depth fix for an AI context bundler that was accidentally including .env files in LLM context.