Rulings
Where specific tools and practices are placed on the scale. Versioned separately from the standard, so a ruling can change without reissuing it.
Nothing here yet
This register is empty, and saying so is more useful than filling it with guesses. Rulings will be published as real questions arrive.
Why it is separate from the standard
Writing tools acquire and lose generative features every few months. A standard that names products is out of date the day it publishes, and reissuing a standard to reclassify one piece of software would be absurd.
So the standard carries the principle — did the tool have to understand your work, and did it write any words that survived — and this page carries the product-level answers. The principle is stable. The answers are not.
What a ruling will look like
A named tool or practice, the number it lands on, and the reasoning in one or two sentences — tied back to the two tests, never to a private judgement about the product.
Where a tool has changed behaviour, the previous ruling stays visible with its date. A declaration made in good faith under an old ruling was made in good faith.
Until then
The two questions in the standard settle most cases without a ruling. Did the tool have to understand your work to do its job? If yes, you are at least a 1. Did it write any of the words that survived? If yes, at least a 4.