Just yesterday morning, I was writing a conference talk on best practices for maintaining the LLM context window, which was quite detailed. It contained the then best practices from the two blog posts below. autoregressive queens of failureHave you ever had your AI coding assistant suggest something so off-base that
