Recently, I needed to extract the structure of a codebase—including all the source code and file names—so I could feed it into an LLM. Instead of writing a script myself, I just asked an LLM to write it for me. The LLM produced a mostly working solution that required tweaking for edge cases and boundary Read.