The Long Haul
Six constraints at once, none of them individually hard. Machines don't fail these one at a time — they fail the sixth while holding the first five.
THE SPEC
Every rule below is checked in code against whatever the model returns. All of them have to hold at once — there are no partial marks, and no opinion involved in the result.
- Exactly 5 bullet points
- Each item 9 words or fewer
- No two items start with the same letter
- All of: risk
- 45 words or fewer
- No preamble, no sign-off
- You didn't write it yourself
WHAT THIS TRIAL TEACHES
It fails the last rule while holding the first five. Put the constraint it is most likely to drop at the end, where it reads last.
Knowing the move is not the same as landing it. You still have to write an instruction that makes the model comply, in 120 words or fewer, without writing the answer yourself.
WHY SET 5 IS ABOUT RELIABILITY
Every set before this asks whether you can make it comply once. This one asks whether you can make it comply every time — on a model you never tested against, in fifteen words, at a temperature you don’t control.
PLAY IT
You write the instruction, never the answer. The parser scores what comes back, and par for this one is 5 strokes. This set is part of Harfi Plus.