Meta Cognitive Reasoning

This is a framework for avoiding the cognitive traps that make LLMs confidently wrong. It enforces evidence before interpretation, multiple hypothesis generation before conclusions, and explicit verification of version claims that might be beyond training data. You'd reach for this when debugging complex issues, conducting code reviews, or any high-stakes analysis where "I assume" or "typically means" could lead you astray. The protocols are opinionated and verbose by design, trading brevity for rigor. Think of it as a checklist system that forces you to show your work, especially when pattern matching wants to jump straight to conclusions.