Not many people will use this, and not many people will know this prompt exists

This prompt is a precision tool for improving any instruction with a single, controlled edit.

You feed it a prompt, and it returns one minimal modification, capped at eight words, that produces the highest improvement in output quality. The change can only be an addition, deletion, or reorder, and it is chosen based on whether it improves ambiguity, clarity, structure, or model behaviour.

What you get back is simple and structured: the proposed change, the updated prompt snippet, and a short explanation grounded in those four criteria. No rewrites, no overthinking, just surgical-level improvement to the weakest part of the instruction.

Use it when you want to tighten prompts without breaking their intent. It is especially useful for refining agent prompts, fixing unclear instructions, or stabilising outputs that feel inconsistent.

What you can expect is small changes with disproportionate impact. It will not reinvent your prompt, it will expose and fix the exact point where it starts to fail.

This is a simple prompt and I mean simple.

```

“Given this prompt, identify the single lowest-effort change (add, delete, or reorder ≤5 words) that produces the largest improvement in output quality.

Explain why that change works.”

```

0 comments