Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Produces a structured plan using the plan genre skeleton: objective, non-negotiables (top 3), time blocks, dependencies, and success criteria. Assembles relevant context before planning. Outputs to the appropriate file (rhythm/week-plan.md for weekly, custom for project-specific).
.claude/skills/miosa-osa-plan/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 130% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -55% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -55% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -56% | 0% |
> Create a structured execution plan with objectives, non-negotiables, and time blocks.
/plan "<objective>" [--horizon <days|week|month>] [--for <person>]Produces a structured plan using the plan genre skeleton: objective, non-negotiables (top 3), time blocks, dependencies, and success criteria. Assembles relevant context before planning. Outputs to the appropriate file (rhythm/week-plan.md for weekly, custom for project-specific).
/assemble relevant topic context.bash# Plan the week /plan "Ship AI Masters pricing page" --horizon week # Plan a project sprint /plan "Platform MVP launch" --horizon month # Plan for a specific person /plan "Bennett's content calendar" --for "Bennett" --horizon week
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 21,107 | 13,520 | -36% | 1 | 1 | 0% | 2,920 | 2,226 | -24% | 0 | 0 | — |
case-02 | pass→pass | 11,497 | 12,210 | +6% | 1 | 1 | 0% | 1,552 | 2,455 | +58% | 0 | 0 | — |
case-03 | pass→pass | 4,908 | 3,325 | -32% | 1 | 1 | 0% | 770 | 760 | -1% | 0 | 0 | — |
case-04 | fail→pass | 11,612 | 6,607 | -43% | 1 | 1 | 0% | 1,754 | 1,249 | -29% | 0 | 0 | — |
case-05 | fail→pass | 1,926 | 2,240 | +16% | 1 | 1 | 0% | 305 | 700 | +130% | 0 | 0 | — |
case-06 | fail→pass | 9,602 | 1,936 | -80% | 1 | 1 | 0% | 1,280 | 578 | -55% | 0 | 0 | — |
case-07 | fail→pass | 8,740 | 1,987 | -77% | 1 | 1 | 0% | 1,372 | 623 | -55% | 0 | 0 | — |
case-08 | fail→pass | 10,960 | 2,410 | -78% | 1 | 1 | 0% | 1,477 | 648 | -56% | 0 | 0 | — |
case-09 | fail→pass | 7,574 | 1,850 | -76% | 1 | 1 | 0% | 1,161 | 558 | -52% | 0 | 0 | — |
case-18 | fail→pass | 7,230 | 1,811 | -75% | 1 | 1 | 0% | 1,022 | 522 | -49% | 0 | 0 | — |
case-10 | pass→pass | 7,039 | 3,352 | -52% | 1 | 1 | 0% | 1,086 | 858 | -21% | 0 | 0 | — |
case-11 | pass→pass | 14,380 | 10,607 | -26% | 1 | 1 | 0% | 2,080 | 1,826 | -12% | 0 | 0 | — |
case-12 | pass→pass | 18,440 | 11,472 | -38% | 1 | 1 | 0% | 2,414 | 1,674 | -31% | 0 | 0 | — |
case-13 | pass→pass | 19,342 | 12,898 | -33% | 1 | 1 | 0% | 2,819 | 2,215 | -21% | 0 | 0 | — |
case-14 | pass→pass | 12,592 | 4,091 | -68% | 1 | 1 | 0% | 1,852 | 1,000 | -46% | 0 | 0 | — |
case-15 | fail→pass | 12,139 | 2,660 | -78% | 1 | 1 | 0% | 1,623 | 790 | -51% | 0 | 0 | — |
case-16 | fail→pass | 11,676 | 3,648 | -69% | 1 | 1 | 0% | 1,820 | 776 | -57% | 0 | 0 | — |
case-17 | fail→pass | 16,871 | 14,956 | -11% | 1 | 1 | 0% | 2,011 | 2,185 | +9% | 0 | 0 | — |
case-19 | fail→fail | 13,832 | 1,731 | -87% | 1 | 1 | 0% | 2,165 | 502 | -77% | 0 | 0 | — |
case-20 | fail→pass | 8,355 | 1,559 | -81% | 1 | 1 | 0% | 1,273 | 525 | -59% | 0 | 0 | — |
case-21 | pass→pass | 20,706 | 21,445 | +4% | 1 | 1 | 0% | 2,652 | 3,285 | +24% | 0 | 0 | — |
case-22 | pass→pass | 10,728 | 9,959 | -7% | 1 | 1 | 0% | 1,473 | 1,463 | -1% | 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 +50 percentage points is the difference between those two pass rates over the 22 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.