Ask HN: LLM agent to firewall vulnerable FOSS libs?

7 points by ATechGuy 3 months ago

As software providers, we use several FOSS libraries. These libraries require periodical security bug fixing, which requires shipping a new release. We are building an LLM agent that can block these security bugs at runtime (like a firewall so to speak) and prevent attackers from exploiting security bugs in vulnerable libraries. Are there any pitfalls that we must avoid? Your feedback is welcome.

wmf 3 months ago

Yeah, the main pitfall is that this has no chance of working.

  • ATechGuy 3 months ago

    Care to elaborate?

    • atlasduo 3 months ago

      I am not the original commenter, but when you are firmly set to work with a hammer, everything starts to look like a nail.

ailef 3 months ago

How would an LLM agent "block these security bugs at runtime" exactly?

  • wmf 3 months ago

    I guess the idea is like spam filtering but for function calls. You could use a deep learning model as a classifier to detect abnormal patterns. You probably wouldn't want to use an LLM though and I'm skeptical about the accuracy and performance.