I feel the same way, so this post inspired me to think about "how to improve intelligence." If I encounter a problem that's difficult to solve (and more automa

I feel the same way, so this post inspired me to think about "how to improve intelligence." If I encounter a problem that's difficult to solve (and more automated tools like Claude Code can't solve it), I might try to package the source code context and send it separately to the LLM chat window. This approach can reduce the length of the context.