> The LLM also can’t really learn. An intuitive response to this problem is to simply continue adding more and more instructions to its pre-prompt, treating that text file as its “memory”, but that just doesn’t work, and probably never will.
The link goes to a "bug report" (really a support request) noting that Claude Code gives a length warning for CLAUDE.md, and asking whether it makes sense to factor out separate files. The first response (several weeks later) is blatantly LLM-generated text running with the idea. It even features:
> Are you sure this works?
> You've asked a fantastic and very insightful question. It gets to the very heart of how CLAUDE.md works, and I apologize if my previous answer was ambiguous.
Painful. Why do people think they are providing a service by doing things like this?
The link goes to a "bug report" (really a support request) noting that Claude Code gives a length warning for CLAUDE.md, and asking whether it makes sense to factor out separate files. The first response (several weeks later) is blatantly LLM-generated text running with the idea. It even features:
> Are you sure this works?
> You've asked a fantastic and very insightful question. It gets to the very heart of how CLAUDE.md works, and I apologize if my previous answer was ambiguous.
Painful. Why do people think they are providing a service by doing things like this?