Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Identify constraints, classify them by type and severity, assess removability, and design removal paths for removable constraints.
.claude/skills/yogsoth-ai-constraint-drilling/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 65% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 6% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 124% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 4% | 0% |
Systematically discover all constraints blocking a candidate, classify them into actionable categories, assess which can be removed, and design concrete removal paths for those that can.
constraint-identification-sop SOP.constraint-classification SOP.removability-assessment SOP for each constraint (parallelizable).removal-path SOP for each removable constraint.| SOP | Stage | Purpose | |-----|-------|---------| | constraint-identification-sop | 1 | Discover constraints using structured methods | | constraint-classification | 2 | Categorize constraints by type | | removability-assessment | 3 | Score removability of each constraint | | removal-path | 4 | Design removal steps and timeline |
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | constraint-classification | Classify constraints into hard constraints, soft constraints, and assumptions. | | constraint-identification-sop | Identify constraints for a candidate using TOC, TRIZ, and Pre-mortem methods. | | removability-assessment | Assess how removable a constraint is with effort estimate and dependency analysis. | | removal-path | Design concrete removal steps for a constraint with timeline and resource needs. |
<!-- END available-tables (generated) -->
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 26,774 | 39,496 | +48% | 1 | 1 | 0% | 4,111 | 6,773 | +65% | 0 | 0 | — |
case-02 | fail→pass | 23,872 | 20,996 | -12% | 1 | 1 | 0% | 3,991 | 3,842 | -4% | 0 | 0 | — |
case-03 | fail→pass | 29,810 | 28,377 | -5% | 1 | 1 | 0% | 4,847 | 5,138 | +6% | 0 | 0 | — |
case-04 | pass→fail | 14,598 | 19,555 | +34% | 1 | 1 | 0% | 2,494 | 3,676 | +47% | 0 | 0 | — |
case-05 | fail→fail | 4,258 | 28,629 | +572% | 1 | 1 | 0% | 729 | 5,620 | +671% | 0 | 0 | — |
case-06 | pass→fail | 15,739 | 33,407 | +112% | 1 | 1 | 0% | 2,350 | 5,695 | +142% | 0 | 0 | — |
case-07 | fail→pass | 14,007 | 26,103 | +86% | 1 | 1 | 0% | 2,041 | 4,581 | +124% | 0 | 0 | — |
case-08 | fail→fail | 9,915 | 31,817 | +221% | 1 | 1 | 0% | 1,444 | 5,795 | +301% | 0 | 0 | — |
case-09 | fail→pass | 23,310 | 19,241 | -17% | 1 | 1 | 0% | 3,608 | 3,759 | +4% | 0 | 0 | — |
case-10 | pass→pass | 11,591 | 3,897 | -66% | 1 | 1 | 0% | 1,629 | 1,199 | -26% | 0 | 0 | — |
case-11 | pass→pass | 12,032 | 22,730 | +89% | 1 | 1 | 0% | 1,678 | 4,216 | +151% | 0 | 0 | — |
case-12 | fail→pass | 16,427 | 12,187 | -26% | 1 | 1 | 0% | 2,425 | 2,406 | -1% | 0 | 0 | — |
case-13 | fail→pass | 3,274 | 1,882 | -43% | 1 | 1 | 0% | 444 | 789 | +78% | 0 | 0 | — |
case-14 | pass→pass | 8,436 | 2,448 | -71% | 1 | 1 | 0% | 1,184 | 927 | -22% | 0 | 0 | — |
case-15 | fail→pass | 10,762 | 1,526 | -86% | 1 | 1 | 0% | 1,514 | 748 | -51% | 0 | 0 | — |
case-16 | fail→pass | 7,764 | 1,683 | -78% | 1 | 1 | 0% | 1,160 | 762 | -34% | 0 | 0 | — |
case-17 | fail→pass | 11,564 | 6,282 | -46% | 1 | 1 | 0% | 1,719 | 1,431 | -17% | 0 | 0 | — |
case-18 | fail→pass | 8,659 | 4,256 | -51% | 1 | 1 | 0% | 1,232 | 1,320 | +7% | 0 | 0 | — |
case-19 | pass→pass | 8,635 | 10,300 | +19% | 1 | 1 | 0% | 1,268 | 2,134 | +68% | 0 | 0 | — |
case-20 | pass→pass | 7,843 | 5,119 | -35% | 1 | 1 | 0% | 1,183 | 1,329 | +12% | 0 | 0 | — |
case-21 | pass→pass | 15,391 | 7,396 | -52% | 1 | 1 | 0% | 2,397 | 1,657 | -31% | 0 | 0 | — |
case-22 | pass→pass | 13,615 | 5,912 | -57% | 1 | 1 | 0% | 1,956 | 1,467 | -25% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +41 percentage points is the difference between those two pass rates over the 22 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.