I’d like to hear more about how you get AI to test what must be a near-infinite combination of possible configurations for software like NGINX. I haven’t looked, but I wonder how you then narrow down those test scenarios to configurations that are more likely to be found in the wild. Is there a survey of NGINX configs out there? How would we know which configurations are more or less likely to be used?
I’d like to hear more about how you get AI to test what must be a near-infinite combination of possible configurations for software like NGINX. I haven’t looked, but I wonder how you then narrow down those test scenarios to configurations that are more likely to be found in the wild. Is there a survey of NGINX configs out there? How would we know which configurations are more or less likely to be used?
This is our workflow:
1. Ask Claude to audit nginx
2. Verify bugs
3. Use Shodan to estimate the popularity of the config
super cool!
any chance you're going to share transcripts' jsonls?