Yes, it seems that you can use #LLM's to reverse #Javascript variable name minification. Working on a tool that uses llama2 locally to make it practical to use on large files (can cost 1000s of dollars on big js bundles if using #ChatGPT ). Dropping a release blog post in a few days for #OSS tool
#bugbounty #infosec
@jehna that’s super interesting. How many passes over the same code do you need to do?
@kellogh llama2 needs one pass per each variable name. ChatGPT is much smarter, I can do ~10 variables per pass since they have the "function calling" feature