You can do this with a modified tool for LLM agents if you really wanted to with an obfuscated prompt/obfuscated tool but what's the point? It is funny the firs

You can do this with a modified tool for LLM agents if you really wanted to with an obfuscated prompt/obfuscated tool but what's the point? It is funny the first time but this is just stick ChatGPT so is unsurprising it would have greater restrictions