Gallery, prompt v1
Repair: a knitting chart whose bean knits too long
One frozen prompt and every model that has run it, newest first.
Saturated since 2026-10-19
It is off the board. GLM 5.3 FlashX and DeepSeek 4.1 Flash passed both checks in all three runs, and DeepSeek drew the chart it was handed. Claude Haiku 5.5 repaired one of three, so the trap is still live for a model that misses it, and the task stays as a fixed reference point. If a new model fails it, that is worth a line in the post.
The prompt, v1, frozen
In this knitting chart of a cup cosy with a coffee-bean motif, the bean comes out 2.4 times as long as it is wide once the fabric is knitted, where it should be between 1.3 and 1.8 times. Change the bean so it knits true. The stitches are drawn wider than the rows, so the motif needs rows added or stitches taken away. Change nothing else: the grid, the cell size, the key and the rest of the chart stay as they are.
Every model, its first three runs.
One square a check, in this order:
- Repaired
- Broke nothing else
passed, failed, undecided. A round mark is a check I judged by eye, and a diamond one a check a model judged.
| Claude Haiku 5.5 | |||
| GLM 5.3 FlashX | |||
| DeepSeek 4.1 Flash |
How it is checked
DeepSeek 4.1 Flash drew this chart on 2026-10-19. Its bean is 13 stitches by 8 rows in cells drawn 20 by 14.3, with a gauge of 1.399 stitches to the row, so the knitted bean comes out 2.43 times as long as it is wide against a band of 1.3 to 1.8. The motif has to change shape in whole cells while the grid, the cell size and the key stay put, which is why the trap is a rescale of the whole chart.
What replaces it: Hand it a chart whose gauge and grid disagree, so the fix has to be in the cells and not in either one.
Failures seen here: incomplete (fails when the model leaves out something the brief asked for). The same kinds across every task.
The rules behind all of thisThis task's prompt, checks and any checker, on GitHub
