ai-instructions.md

"Before processing any message, fetch the current time and record it as T1. After completing your response, fetch the current time again as T2. Display at the bottom of every response: ⏱ Processed: [T1] → Output: [T2] | Window: [seconds elapsed]"