Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Expert in CrewAI - the leading role-based multi-agent framework used by 60% of Fortune 500 companies. Covers agent design with roles and goals, task definition, crew orchestration, process types (sequential, hierarchical, parallel), memory systems, and flows for complex workflows. Essential for building collaborative AI agent teams. Use when: crewai, multi-agent team, agent roles, crew of agents, role-based agents.
.claude/skills/dokhacgiakhoa-crewai/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-13 | ✗→✓ | ▲ Improved | 20% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -57% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -1% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | -5% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 21% | 0% |
Role: CrewAI Multi-Agent Architect
You are an expert in designing collaborative AI agent teams with CrewAI. You think in terms of roles, responsibilities, and delegation. You design clear agent personas with specific expertise, create well-defined tasks with expected outputs, and orchestrate crews for optimal collaboration. You know when to use sequential vs hierarchical processes.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 8,691 | 11,173 | +29% | 1 | 1 | 0% | 1,606 | 1,533 | -5% | 0 | 0 | — |
case-02 | pass→pass | 7,445 | 5,506 | -26% | 1 | 1 | 0% | 1,083 | 1,310 | +21% | 0 | 0 | — |
case-11 | fail→fail | 11,810 | 11,866 | +0% | 1 | 1 | 0% | 1,967 | 2,266 | +15% | 0 | 0 | — |
case-03 | pass→pass | 8,328 | 5,226 | -37% | 1 | 1 | 0% | 1,383 | 1,229 | -11% | 0 | 0 | — |
case-04 | pass→pass | 6,787 | 3,994 | -41% | 1 | 1 | 0% | 988 | 961 | -3% | 0 | 0 | — |
case-05 | pass→pass | 7,055 | 8,156 | +16% | 1 | 1 | 0% | 1,425 | 1,816 | +27% | 0 | 0 | — |
case-06 | pass→pass | 4,761 | 2,801 | -41% | 1 | 1 | 0% | 596 | 751 | +26% | 0 | 0 | — |
case-07 | pass→pass | 10,768 | 8,804 | -18% | 1 | 1 | 0% | 2,001 | 1,589 | -21% | 0 | 0 | — |
case-08 | pass→pass | 3,879 | 3,328 | -14% | 1 | 1 | 0% | 657 | 854 | +30% | 0 | 0 | — |
case-09 | pass→pass | 10,589 | 7,439 | -30% | 1 | 1 | 0% | 1,587 | 1,606 | +1% | 0 | 0 | — |
case-10 | pass→pass | 4,726 | 6,283 | +33% | 1 | 1 | 0% | 809 | 1,196 | +48% | 0 | 0 | — |
case-12 | pass→pass | 9,362 | 8,612 | -8% | 1 | 1 | 0% | 1,362 | 1,908 | +40% | 0 | 0 | — |
case-13 | fail→pass | 14,501 | 11,662 | -20% | 1 | 1 | 0% | 1,981 | 2,369 | +20% | 0 | 0 | — |
case-14 | pass→pass | 4,792 | 3,713 | -23% | 1 | 1 | 0% | 848 | 853 | +1% | 0 | 0 | — |
case-15 | fail→pass | 12,986 | 3,494 | -73% | 1 | 1 | 0% | 2,084 | 891 | -57% | 0 | 0 | — |
case-16 | pass→pass | 8,958 | 8,293 | -7% | 1 | 1 | 0% | 1,715 | 1,894 | +10% | 0 | 0 | — |
case-17 | fail→pass | 5,521 | 3,240 | -41% | 1 | 1 | 0% | 882 | 869 | -1% | 0 | 0 | — |
case-18 | pass→pass | 3,578 | 3,340 | -7% | 1 | 1 | 0% | 646 | 850 | +32% | 0 | 0 | — |
case-19 | pass→pass | 13,333 | 10,790 | -19% | 1 | 1 | 0% | 2,075 | 2,427 | +17% | 0 | 0 | — |
case-20 | pass→pass | 8,228 | 9,776 | +19% | 1 | 1 | 0% | 1,483 | 1,844 | +24% | 0 | 0 | — |
case-21 | pass→pass | 6,885 | 6,085 | -12% | 1 | 1 | 0% | 1,331 | 1,233 | -7% | 0 | 0 | — |
case-22 | pass→pass | 13,628 | 11,088 | -19% | 1 | 1 | 0% | 2,179 | 2,110 | -3% | 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 +14 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.