Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Executes the agent's primary task. Takes user input, processes it according to the
.claude/skills/miosa-osa-do/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 55% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -42% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 54% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 323% | 0% |
> The agent's primary skill. Customize this to match your agent's purpose.
/do <input> [--format <format>] [--depth <brief|detailed>]Executes the agent's primary task. Takes user input, processes it according to the agent's domain expertise, and produces structured output.
Replace this description with your agent's specific capability:
| Param | Default | Options | Description | |-------|---------|---------|-------------| | --format | markdown | markdown, json, text | Output format | | --depth | brief | brief, detailed | Level of detail |
bash# Basic usage /do "Review this function for bugs" # Detailed output /do "Summarize this meeting transcript" --depth detailed # JSON output for programmatic use /do "Extract entities from this text" --format json
markdown## Result ### Summary [Brief answer to the request] ### Details [Structured breakdown — customize per domain] ### Confidence [High | Medium | Low] — [reason if not High]
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-12 | fail→fail | 46,831 | 15,631 | -67% | 1 | 1 | 0% | 6,865 | 3,144 | -54% | 0 | 0 | — |
case-01 | pass→fail | 2,817 | 27,322 | +870% | 1 | 1 | 0% | 454 | 1,540 | +239% | 0 | 0 | — |
case-02 | fail→pass | 17,535 | 49,590 | +183% | 1 | 1 | 0% | 875 | 1,358 | +55% | 0 | 0 | — |
case-03 | fail→pass | 11,866 | 7,499 | -37% | 1 | 1 | 0% | 1,815 | 1,781 | -2% | 0 | 0 | — |
case-04 | pass→pass | 2,713 | 5,575 | +105% | 1 | 1 | 0% | 341 | 1,080 | +217% | 0 | 0 | — |
case-05 | pass→pass | 21,805 | 12,669 | -42% | 1 | 1 | 0% | 4,649 | 2,451 | -47% | 0 | 0 | — |
case-06 | fail→pass | 12,079 | 4,941 | -59% | 1 | 1 | 0% | 1,713 | 992 | -42% | 0 | 0 | — |
case-07 | pass→pass | 2,339 | 6,262 | +168% | 1 | 1 | 0% | 338 | 1,219 | +261% | 0 | 0 | — |
case-08 | fail→fail | 42,796 | 33,359 | -22% | 1 | 1 | 0% | 1,814 | 1,558 | -14% | 0 | 0 | — |
case-09 | fail→pass | 80,110 | 15,817 | -80% | 1 | 1 | 0% | 1,079 | 1,662 | +54% | 0 | 0 | — |
case-10 | fail→pass | 2,813 | 9,126 | +224% | 1 | 1 | 0% | 411 | 1,738 | +323% | 0 | 0 | — |
case-11 | fail→fail | 7,867 | 9,651 | +23% | 1 | 1 | 0% | 1,275 | 1,966 | +54% | 0 | 0 | — |
case-13 | pass→pass | 4,203 | 9,970 | +137% | 1 | 1 | 0% | 590 | 1,740 | +195% | 0 | 0 | — |
case-14 | fail→pass | 36,440 | 5,907 | -84% | 1 | 1 | 0% | 2,687 | 1,432 | -47% | 0 | 0 | — |
case-15 | fail→pass | 6,321 | 5,867 | -7% | 1 | 1 | 0% | 922 | 1,224 | +33% | 0 | 0 | — |
case-16 | fail→pass | 5,915 | 4,717 | -20% | 1 | 1 | 0% | 827 | 1,180 | +43% | 0 | 0 | — |
case-17 | fail→pass | 9,352 | 15,512 | +66% | 1 | 1 | 0% | 1,793 | 2,524 | +41% | 0 | 0 | — |
case-18 | pass→pass | 13,100 | 8,769 | -33% | 1 | 1 | 0% | 2,155 | 1,784 | -17% | 0 | 0 | — |
case-19 | fail→pass | 3,098 | 5,962 | +92% | 1 | 1 | 0% | 455 | 1,141 | +151% | 0 | 0 | — |
case-20 | fail→pass | 4,778 | 4,217 | -12% | 1 | 1 | 0% | 507 | 901 | +78% | 0 | 0 | — |
case-21 | pass→pass | 12,593 | 10,854 | -14% | 1 | 1 | 0% | 2,431 | 2,450 | +1% | 0 | 0 | — |
case-22 | fail→pass | 7,491 | 7,218 | -4% | 1 | 1 | 0% | 1,664 | 1,402 | -16% | 0 | 0 | — |
case-23 | fail→pass | 16,860 | 10,496 | -38% | 1 | 1 | 0% | 2,461 | 1,781 | -28% | 0 | 0 | — |
case-24 | pass→pass | 2,898 | 4,508 | +56% | 1 | 1 | 0% | 371 | 1,006 | +171% | 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 +50 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.
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.