Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Decompose system into functional components, identify dependencies, and surface trimming candidates.
.claude/skills/yogsoth-ai-component-decomposition/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -5% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 0% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 65% | 0% |
Decompose a system into its functional components, map dependencies, and identify candidates for trimming or surgical modification.
Build complete functional model using function-model-construction SOP. Identify all components and their interactions (useful, harmful, insufficient, excessive).
For each component, identify key parameters using parameter-identification SOP. Map which parameters are shared, conflicting, or independent.
Evaluate each component for trimming potential via trimming-execution SOP criteria: high harmful-function ratio, function redistributable to neighbors, low integration cost.
| Metric | Floor | |--------|-------| | Components identified | ≥5 | | Interactions mapped | ≥8 | | Trimming candidates | ≥3 | | Parameters per component | ≥2 |
| SOP | Role | |-----|------| | function-model-construction | Stage 1 — build functional model | | parameter-identification | Stage 2 — extract component parameters | | trimming-execution | Stage 3 — evaluate trimming feasibility | | surgery-operation | Post — apply surgical operations to candidates |
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | function-model-construction | Build substance-field functional model of a system, annotating useful, harmful, insufficient, and excessive interactions. | | trimming-execution | Progressively remove components from a system while verifying function preservation through redistribution. |
<!-- 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 | 21,936 | 17,118 | -22% | 1 | 1 | 0% | 3,492 | 3,315 | -5% | 0 | 0 | — |
case-02 | fail→fail | 38,629 | 29,086 | -25% | 1 | 1 | 0% | 6,211 | 5,196 | -16% | 0 | 0 | — |
case-03 | fail→pass | 25,576 | 20,287 | -21% | 1 | 1 | 0% | 4,481 | 3,920 | -13% | 0 | 0 | — |
case-04 | pass→fail | 14,970 | 22,181 | +48% | 1 | 1 | 0% | 2,331 | 4,182 | +79% | 0 | 0 | — |
case-05 | pass→fail | 23,126 | 34,621 | +50% | 1 | 1 | 0% | 4,132 | 6,553 | +59% | 0 | 0 | — |
case-06 | pass→pass | 22,086 | 32,535 | +47% | 1 | 1 | 0% | 3,233 | 5,053 | +56% | 0 | 0 | — |
case-07 | fail→pass | 25,974 | 22,912 | -12% | 1 | 1 | 0% | 3,938 | 3,957 | +0% | 0 | 0 | — |
case-08 | fail→pass | 25,707 | 22,277 | -13% | 1 | 1 | 0% | 3,601 | 3,839 | +7% | 0 | 0 | — |
case-09 | fail→pass | 20,782 | 26,977 | +30% | 1 | 1 | 0% | 3,010 | 4,968 | +65% | 0 | 0 | — |
case-10 | fail→fail | 21,943 | 29,143 | +33% | 1 | 1 | 0% | 3,414 | 5,048 | +48% | 0 | 0 | — |
case-11 | fail→pass | 24,902 | 22,622 | -9% | 1 | 1 | 0% | 3,889 | 5,102 | +31% | 0 | 0 | — |
case-12 | fail→pass | 23,656 | 17,909 | -24% | 1 | 1 | 0% | 3,624 | 3,202 | -12% | 0 | 0 | — |
case-13 | fail→pass | 27,381 | 19,412 | -29% | 1 | 1 | 0% | 4,311 | 3,571 | -17% | 0 | 0 | — |
case-14 | fail→pass | 25,711 | 21,403 | -17% | 1 | 1 | 0% | 3,952 | 3,901 | -1% | 0 | 0 | — |
case-15 | fail→pass | 21,653 | 25,068 | +16% | 1 | 1 | 0% | 3,934 | 4,444 | +13% | 0 | 0 | — |
case-16 | fail→pass | 24,934 | 26,646 | +7% | 1 | 1 | 0% | 3,683 | 4,588 | +25% | 0 | 0 | — |
case-17 | fail→pass | 24,249 | 28,023 | +16% | 1 | 1 | 0% | 3,900 | 5,070 | +30% | 0 | 0 | — |
case-18 | fail→pass | 19,767 | 20,381 | +3% | 1 | 1 | 0% | 3,064 | 3,837 | +25% | 0 | 0 | — |
case-19 | fail→fail | 18,832 | 17,414 | -8% | 1 | 1 | 0% | 2,770 | 3,307 | +19% | 0 | 0 | — |
case-20 | fail→pass | 27,125 | 19,270 | -29% | 1 | 1 | 0% | 4,073 | 3,488 | -14% | 0 | 0 | — |
case-21 | fail→pass | 23,563 | 26,480 | +12% | 1 | 1 | 0% | 3,384 | 4,633 | +37% | 0 | 0 | — |
case-22 | fail→fail | 22,935 | 36,666 | +60% | 1 | 1 | 0% | 3,494 | 6,236 | +78% | 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 +59 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.