Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Contextual help for OptimalOS commands and capabilities.
.claude/skills/miosa-osa-help/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | -57% | 0% |
| case-01 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -37% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -64% | 0% |
Contextual help for OptimalOS commands and capabilities.
Shows available skills, agents, and engine commands. Context-aware: suggests relevant commands based on current cognitive mode and recent activity.
/help # Show all available skills
/help search # Detail on /search skill
/help agents # List available agents
/help modes # Explain cognitive modes
/help genres # List available genres| Skill | Description | |-------|-------------| | /ingest | Ingest signals into knowledge base | | /search | Hybrid search across all contexts | | /assemble | Build tiered context bundle | | /health | Run system diagnostics | | /reweave | Find stale contexts, suggest updates | | /remember | Capture friction patterns | | /verify | Test L0 fidelity | | /graph | Analyze knowledge graph | | /reflect | Find missing entity relationships | | /rethink | Evidence synthesis | | /simulate | Monte Carlo scenario planning | | /setup | System setup and configuration | | /boot | Daily boot sequence |
| Agent | Role | |-------|------| | knowledge-guide | Genre + routing guidance | | signal-processor | Full intake pipeline | | context-assembler | Tiered context loading | | health-monitor | System diagnostics + maintenance |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | fail→fail | 12,862 | 8,230 | -36% | 1 | 1 | 0% | 1,704 | 1,619 | -5% | 0 | 0 | — |
case-07 | fail→pass | 11,968 | 3,034 | -75% | 1 | 1 | 0% | 1,839 | 786 | -57% | 0 | 0 | — |
case-01 | fail→pass | 11,757 | 9,817 | -17% | 1 | 1 | 0% | 1,965 | 1,581 | -20% | 0 | 0 | — |
case-03 | fail→pass | 7,442 | 2,713 | -64% | 1 | 1 | 0% | 961 | 686 | -29% | 0 | 0 | — |
case-04 | fail→pass | 9,802 | 3,196 | -67% | 1 | 1 | 0% | 1,075 | 680 | -37% | 0 | 0 | — |
case-05 | fail→pass | 13,766 | 2,891 | -79% | 1 | 1 | 0% | 1,878 | 675 | -64% | 0 | 0 | — |
case-06 | fail→pass | 11,197 | 1,817 | -84% | 1 | 1 | 0% | 1,641 | 539 | -67% | 0 | 0 | — |
case-08 | fail→pass | 12,041 | 2,398 | -80% | 1 | 1 | 0% | 1,836 | 557 | -70% | 0 | 0 | — |
case-09 | fail→pass | 12,662 | 2,846 | -78% | 1 | 1 | 0% | 1,730 | 673 | -61% | 0 | 0 | — |
case-10 | fail→pass | 11,956 | 2,695 | -77% | 1 | 1 | 0% | 1,744 | 728 | -58% | 0 | 0 | — |
case-11 | fail→pass | 14,003 | 2,630 | -81% | 1 | 1 | 0% | 2,114 | 687 | -68% | 0 | 0 | — |
case-12 | fail→pass | 11,065 | 3,120 | -72% | 1 | 1 | 0% | 1,747 | 784 | -55% | 0 | 0 | — |
case-13 | fail→pass | 17,811 | 2,453 | -86% | 1 | 1 | 0% | 2,588 | 649 | -75% | 0 | 0 | — |
case-14 | fail→pass | 7,979 | 2,829 | -65% | 1 | 1 | 0% | 1,330 | 643 | -52% | 0 | 0 | — |
case-15 | fail→pass | 5,477 | 2,739 | -50% | 1 | 1 | 0% | 834 | 627 | -25% | 0 | 0 | — |
case-16 | fail→pass | 10,086 | 2,143 | -79% | 1 | 1 | 0% | 1,484 | 592 | -60% | 0 | 0 | — |
case-17 | fail→pass | 13,396 | 1,714 | -87% | 1 | 1 | 0% | 1,846 | 547 | -70% | 0 | 0 | — |
case-18 | fail→pass | 10,255 | 2,039 | -80% | 1 | 1 | 0% | 1,662 | 583 | -65% | 0 | 0 | — |
case-19 | pass→pass | 9,539 | 1,932 | -80% | 1 | 1 | 0% | 1,488 | 630 | -58% | 0 | 0 | — |
case-20 | pass→pass | 6,119 | 9,893 | +62% | 1 | 1 | 0% | 781 | 1,891 | +142% | 0 | 0 | — |
case-21 | pass→pass | 12,311 | 8,432 | -32% | 1 | 1 | 0% | 2,190 | 1,712 | -22% | 0 | 0 | — |
case-22 | pass→pass | 15,039 | 11,077 | -26% | 1 | 1 | 0% | 2,685 | 2,370 | -12% | 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 +77 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.