I would honestly hope that the LLM usage would help with that very thing. If AI systems automatically perform security audits on brand-new NPM releases, doesn't that accomplish a similar purpose?