Reasoning Summary
A user- or developer-facing summary of a model's reasoning process without exposing the raw hidden chain of thought.
Plain English
It shows the useful shape of the model's thinking without dumping every internal token.
Example
A debugging agent returns a short summary of why it chose the suspected failing module before showing the patch.
Why it matters
Reasoning summaries are becoming a practical compromise between transparency, safety, and implementation detail.