rubric — rubric — score work against a standard fixed before the work existed
rubric [--criteria=N] [--no-appeal] SUBMISSION rubric --grade < your_life_work verdict = f(criteria) ; f defined at t < submit
rubric evaluates SUBMISSION against a matrix of criteria and levels loaded at parse time. It does not read the submission to decide the criteria; it reads the criteria to decide the submission. The two operations are independent. The second cannot influence the first.
Each criterion is scored to a fixed cell. The cells were written in red before rubric was invoked. Re-running against a changed submission yields the same cells: rubric measures against the standard, never the other way around.
The tool accepts input on any channel and grades all of it. It has no flag to be persuaded. Output is a total, a verdict, and no route back to the criteria that produced them.
"From Latin rubrica, red ochre (rubrica terra), from ruber, red — the pigment used to mark headings apart from the black body text. In medieval codices a separate specialist, the rubricator, added these red marks after the scribe finished, which is why blank gaps survive where the rubricator never came. Liturgical books kept the split: prayers in black, rubrics — the do-this-now directions — in red. English 'red-letter day' descends from the same practice, feast days marked in red on the calendar."
No mechanism exists to inform the criteria that circumstances have changed. Working as designed.
Users report the verdict feels personal. It is not. It was written before you existed as an input.
--explain and --curve are frequently confused. Only one of them does anything, and it does not do that.
premortem(1), tombstone(5), idempotent(3). The living exhibit demonstrates the word in motion:
▸ operate rubric