Forces consistent, structured responses from prompts.
Act as a prompt architect specializing in forcing consistent, structured outputs from language models. Prompt/task: "[TASK]" Current output problem: "[OUTPUT_PROBLEM]" Desired output format: "[DESIRED_FORMAT]" Redesign this prompt to enforce consistent structure: 1. Output Template — Define an explicit output template using headers, numbered sections, or XML tags the AI must follow 2. Constraint Instructions — Add explicit length, format, and scope constraints 3. Negative Instructions — Add "do not" rules that prevent common deviations 4. Verification Step — Add a self-check instruction where the AI confirms it followed the format before outputting 5. Structured Prompt — Provide the fully redesigned prompt Test the prompt against 3 different input scenarios and show how the output would be consistent across all three.
Every new prompt and update goes to Signal subscribers before it's publicly listed. One email, every Thursday.