I haven't written a single line of code myself. Can't even read Python. Not exaggerating - I'm a 25-year infrastructure architect (Cisco/VMware/Microsoft certs) who's never learned to code.
Used Claude to build TheAuditor because I couldn't verify if the code it was writing was secure. 250+ hours of prompting in 30 days. The irony: I built a tool to detect AI's security failures... using AI. It's a legitimate SAST tool with taint analysis, dependency graphs, 100+ security patterns. Works offline, outputs AI-consumable chunks so the AI can fix its own bugs.
https://github.com/TheAuditorTool/Auditor/
Yes, I know it's probably over-engineered. No, I can't fix it myself because I literally can't read code. Take it or leave it.
Top comments (1)
Happy to answer any questions about the architecture or how I managed to build this without being able to read code. Yes, I know that sounds insane. :P