rogee/superpowers
1
0
Fork 0
mirror of https://github.com/obra/superpowers.git synced 2026-06-13 06:09:04 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix(writing-skills): hang backfire mechanism on the separated prohibition-vs-recipe comparison (NEW-4); control comparison stated as trend

Browse Source
This commit is contained in:
Jesse Vincent
2026-06-11 11:30:31 -07:00
parent 391c33edf8
commit 9887c17b31
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
skills/writing-skills/SKILL.md
View File

@@ -467,7 +467,7 @@ Before writing guidance, classify the baseline failure. The form that bulletproo
| Omits a required element from something they already produce | Structural: REQUIRED field or slot in the template they fill in | Prose reminders near the template | | Omits a required element from something they already produce | Structural: REQUIRED field or slot in the template they fill in | Prose reminders near the template |
| Behavior should depend on a condition | Conditional keyed to an observable predicate ("if the brief exists, reference it") | Unconditional rule + exemption clauses | | Behavior should depend on a condition | Conditional keyed to an observable predicate ("if the brief exists, reference it") | Unconditional rule + exemption clauses |
**Why prohibitions backfire on shaping problems:** under a competing incentive ("make the prompt self-contained"), agents negotiate with "don't X" and produce MORE of the unwanted content than with no guidance at all. In head-to-head wording tests on dispatch-prompt guidance, the prohibition arm scored worse than the no-guidance control while the recipe arm scored best — micro-test your own case rather than assuming, but never reach for the prohibition by default. A recipe leaves nothing to negotiate: the output matches the stated shape or it doesn't. **Why prohibitions backfire on shaping problems:** under a competing incentive ("make the prompt self-contained"), agents negotiate with "don't X". In head-to-head wording tests on dispatch-prompt guidance, the prohibition arm produced clearly more of the unwanted content than the recipe arm (fully separated distributions), and trended worse than even the no-guidance control — micro-test your own case rather than assuming, but never reach for the prohibition by default. A recipe leaves nothing to negotiate: the output matches the stated shape or it doesn't.
**Rules for whichever form you pick:** **Rules for whichever form you pick:**
- **No nuance clauses.** "Don't X unless it matters" reopens the negotiation — appending a single nuance clause to a winning recipe degraded it from consistent to noisy in the same wording tests. Express a real exception as its own conditional on an observable predicate. - **No nuance clauses.** "Don't X unless it matters" reopens the negotiation — appending a single nuance clause to a winning recipe degraded it from consistent to noisy in the same wording tests. Express a real exception as its own conditional on an observable predicate.