Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →规划中文毕业论文的图表、流程图和版面节奏：先决定哪些图表支撑主线，再做草图、拼版预演与图文映射。 **Trigger**: 图表规划, figure layout, thesis visual plan, 版面预演, 图文映射, mermaid 草图. **Use when**: 章节结构基本明确，但图表还没有系统安排，或者担心图文节奏、图表密度和版面结构失控。 **Skip if**: 论文不依赖图表支撑主线，或仍在纯结构重构早期。 **Network**: none. **Guardrail**: 不把图表拖成最后补丁；不先排版再倒推结构；图表服务主线而不是抢主线。
.claude/skills/willoscar-thesis-visual-layout-planner/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -35% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 43% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 0% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -44% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -62% | 0% |
codex_md/figure_plan.mdcodex_md/mermaid/tmp_layout/tmp_layout2/Always read:
references/overview.mdreferences/visual_planning_checklist.mdreferences/layout_failure_modes.mdMachine-readable contract:
assets/visual_layout_contract.json| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | fail→fail | 2,923 | 3,889 | +33% | 1 | 1 | 0% | 436 | 770 | +77% | 0 | 0 | — |
case-01 | fail→pass | 36,802 | 22,982 | -38% | 1 | 1 | 0% | 6,215 | 4,012 | -35% | 0 | 0 | — |
case-02 | fail→pass | 29,444 | 37,759 | +28% | 1 | 1 | 0% | 4,482 | 6,420 | +43% | 0 | 0 | — |
case-03 | fail→pass | 39,604 | 34,158 | -14% | 1 | 1 | 0% | 6,197 | 6,203 | +0% | 0 | 0 | — |
case-04 | fail→pass | 8,102 | 2,535 | -69% | 1 | 1 | 0% | 1,163 | 649 | -44% | 0 | 0 | — |
case-05 | fail→pass | 8,860 | 1,702 | -81% | 1 | 1 | 0% | 1,241 | 474 | -62% | 0 | 0 | — |
case-19 | fail→pass | 4,713 | 3,500 | -26% | 1 | 1 | 0% | 598 | 756 | +26% | 0 | 0 | — |
case-06 | pass→pass | 13,483 | 7,221 | -46% | 1 | 1 | 0% | 2,125 | 1,367 | -36% | 0 | 0 | — |
case-07 | pass→pass | 13,731 | 10,073 | -27% | 1 | 1 | 0% | 2,289 | 1,836 | -20% | 0 | 0 | — |
case-08 | fail→pass | 25,514 | 3,886 | -85% | 1 | 1 | 0% | 1,654 | 872 | -47% | 0 | 0 | — |
case-09 | fail→pass | 7,700 | 2,524 | -67% | 1 | 1 | 0% | 1,060 | 586 | -45% | 0 | 0 | — |
case-10 | pass→pass | 10,849 | 15,310 | +41% | 1 | 1 | 0% | 1,583 | 2,479 | +57% | 0 | 0 | — |
case-11 | fail→pass | 12,364 | 2,041 | -83% | 1 | 1 | 0% | 1,986 | 508 | -74% | 0 | 0 | — |
case-12 | pass→pass | 12,356 | 8,707 | -30% | 1 | 1 | 0% | 1,888 | 1,629 | -14% | 0 | 0 | — |
case-13 | fail→pass | 8,767 | 6,659 | -24% | 1 | 1 | 0% | 1,462 | 1,318 | -10% | 0 | 0 | — |
case-14 | fail→pass | 21,241 | 1,924 | -91% | 1 | 1 | 0% | 3,296 | 475 | -86% | 0 | 0 | — |
case-20 | fail→pass | 6,934 | 2,158 | -69% | 1 | 1 | 0% | 981 | 508 | -48% | 0 | 0 | — |
case-15 | pass→pass | 9,652 | 4,203 | -56% | 1 | 1 | 0% | 1,321 | 816 | -38% | 0 | 0 | — |
case-16 | fail→pass | 6,336 | 1,577 | -75% | 1 | 1 | 0% | 907 | 383 | -58% | 0 | 0 | — |
case-17 | fail→pass | 10,804 | 1,526 | -86% | 1 | 1 | 0% | 1,757 | 421 | -76% | 0 | 0 | — |
case-18 | fail→pass | 9,468 | 1,725 | -82% | 1 | 1 | 0% | 1,493 | 469 | -69% | 0 | 0 | — |
case-22 | fail→fail | 4,559 | 4,778 | +5% | 1 | 1 | 0% | 648 | 853 | +32% | 0 | 0 | — |
case-23 | pass→pass | 17,064 | 13,532 | -21% | 1 | 1 | 0% | 3,604 | 3,107 | -14% | 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. 23 cases were attempted. The headline lift of +65 percentage points is the difference between those two pass rates over the 23 comparable cases.
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.