I think it's funny how goblin mode this whole hack is. The memory scanner itself was clearly written by an LLM (using python??) and the way this person goes about hacking the game is very non-reverse engineer but instead someone equipped with very capable tools. No shade to the dude to be clear, I think it's sort of incredibly how possible this stuff is now due to LLMs and doesn't require someone to know how to use Ghidra.
I've seen this argument plenty of times with respect to LLMs writing code, but this is the first time I've seen someone roll it out for using an LLM to answer questions about code that is being fed into it as input!
I was intrigued as to how it would intercept a conversation and then pause the game for long enough for the LLM to return a response, so I used https://gitingest.com/vuciv/animal-crossing-llm-mod to dump the 40,000 tokens into Claude Opus 4.1 and asked it: https://claude.ai/share/66c52dc8-9ebd-4db7-8159-8f694e06b381
The trick is the watch_dialogue() function which polls every 0.1 seconds and then answers with placeholder text: https://github.com/vuciv/animal-crossing-llm-mod/blob/cc9b6b...
So the user gets a "press A to continue" button and hopefully the LLM has finished by the time they press that button.