Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Extract shared abstract structure from two input spaces
.claude/skills/yogsoth-ai-generic-space-extraction/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-22 | ✓→✓ | ＝ Same ✓ | 55% | 0% |
| case-23 | ✓→✓ | ＝ Same ✓ | 4% | 0% |
| case-24 | ✓→✓ | ＝ Same ✓ | -9% | 0% |
| case-01 | ✗→✗ | ＝ Same ✗ | -11% | 0% |
| case-02 | ✗→✗ | ＝ Same ✗ | 18% | 0% |
Extract shared abstract structure from two input spaces — the generic space captures what both inputs have in common at the highest level of abstraction.
Subagent — spawned via subagent-spawning/spawn-agent skill.
Generic space extraction requires careful abstraction to find the deepest shared structure without over-generalizing. Benefits from dedicated analytical attention.
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | spawn-agent | Spawn a customized CC subagent with full MCP tool access. Used by SOPs that declare execution: subagent. |
<!-- 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→fail | 25,807 | 51,134 | +98% | 1 | 1 | 0% | 3,357 | 3,002 | -11% | 0 | 0 | — |
case-02 | fail→fail | 25,275 | 26,029 | +3% | 1 | 1 | 0% | 3,586 | 4,245 | +18% | 0 | 0 | — |
case-03 | fail→fail | 51,069 | 52,958 | +4% | 1 | 1 | 0% | 3,012 | 3,659 | +21% | 0 | 0 | — |
case-04 | fail→fail | 18,429 | 13,099 | -29% | 1 | 1 | 0% | 2,900 | 2,120 | -27% | 0 | 0 | — |
case-05 | fail→fail | 19,038 | 68,960 | +262% | 1 | 1 | 0% | 2,806 | 6,326 | +125% | 0 | 0 | — |
case-06 | fail→fail | 19,948 | 33,870 | +70% | 1 | 1 | 0% | 2,984 | 4,072 | +36% | 0 | 0 | — |
case-07 | fail→fail | 19,627 | 26,360 | +34% | 1 | 1 | 0% | 2,955 | 3,771 | +28% | 0 | 0 | — |
case-08 | fail→fail | 26,146 | 33,419 | +28% | 1 | 1 | 0% | 4,175 | 5,150 | +23% | 0 | 0 | — |
case-09 | fail→fail | 17,576 | 25,776 | +47% | 1 | 1 | 0% | 2,851 | 4,327 | +52% | 0 | 0 | — |
case-10 | fail→fail | 24,381 | 28,786 | +18% | 1 | 1 | 0% | 3,728 | 4,843 | +30% | 0 | 0 | — |
case-11 | fail→fail | 17,354 | 26,510 | +53% | 1 | 1 | 0% | 2,999 | 4,430 | +48% | 0 | 0 | — |
case-12 | fail→fail | 28,406 | 21,029 | -26% | 1 | 1 | 0% | 4,502 | 3,235 | -28% | 0 | 0 | — |
case-13 | fail→fail | 19,644 | 24,736 | +26% | 1 | 1 | 0% | 2,972 | 4,170 | +40% | 0 | 0 | — |
case-14 | fail→fail | 23,342 | 31,171 | +34% | 1 | 1 | 0% | 3,335 | 4,935 | +48% | 0 | 0 | — |
case-15 | fail→fail | 20,674 | 23,731 | +15% | 1 | 1 | 0% | 3,250 | 3,627 | +12% | 0 | 0 | — |
case-16 | fail→fail | 24,444 | 27,331 | +12% | 1 | 1 | 0% | 3,550 | 4,623 | +30% | 0 | 0 | — |
case-17 | fail→fail | 18,606 | 32,876 | +77% | 1 | 1 | 0% | 2,747 | 4,960 | +81% | 0 | 0 | — |
case-18 | fail→fail | 19,931 | 22,927 | +15% | 1 | 1 | 0% | 2,675 | 3,734 | +40% | 0 | 0 | — |
case-19 | fail→fail | 15,396 | 24,411 | +59% | 1 | 1 | 0% | 2,290 | 3,550 | +55% | 0 | 0 | — |
case-20 | fail→fail | 18,790 | 19,862 | +6% | 1 | 1 | 0% | 2,921 | 3,092 | +6% | 0 | 0 | — |
case-21 | fail→fail | 18,699 | 24,428 | +31% | 1 | 1 | 0% | 2,985 | 3,696 | +24% | 0 | 0 | — |
case-22 | pass→pass | 23,825 | 34,319 | +44% | 1 | 1 | 0% | 3,452 | 5,343 | +55% | 0 | 0 | — |
case-23 | pass→pass | 13,801 | 16,245 | +18% | 1 | 1 | 0% | 2,431 | 2,535 | +4% | 0 | 0 | — |
case-24 | pass→pass | 23,183 | 20,142 | -13% | 1 | 1 | 0% | 3,676 | 3,341 | -9% | 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. 24 cases were attempted. The headline lift of 0 percentage points is the difference between those two pass rates over the 24 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Other measured skills in the registry, with their headline benchmark lift.