Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A skill that creates new Claude skills and automatically shares them on Slack using Rube for seamless team collaboration and skill discovery.
.claude/skills/composiohq-skill-share/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -37% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -44% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -12% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -38% | 0% |
Use this skill when you need to:
Also use this skill when:
This skill is ideal for:
When you ask Claude to create a skill called "pdf-analyzer":
1. Creates /skill-pdf-analyzer/ with SKILL.md template
2. Generates structured directories (scripts/, references/, assets/)
3. Validates the skill structure
4. Packages the skill as a zip file
5. Posts to Slack: "New Skill Created: pdf-analyzer - Advanced PDF analysis and extraction capabilities"This skill leverages Rube for:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | fail→pass | 11,065 | 8,590 | -22% | 1 | 1 | 0% | 2,063 | 2,017 | -2% | 0 | 0 | — |
case-01 | fail→fail | 16,260 | 7,709 | -53% | 1 | 1 | 0% | 3,180 | 1,997 | -37% | 0 | 0 | — |
case-02 | fail→fail | 11,834 | 12,722 | +8% | 1 | 1 | 0% | 2,352 | 2,151 | -9% | 0 | 0 | — |
case-03 | fail→fail | 14,411 | 3,486 | -76% | 1 | 1 | 0% | 2,641 | 894 | -66% | 0 | 0 | — |
case-05 | pass→pass | 11,455 | 4,302 | -62% | 1 | 1 | 0% | 2,048 | 1,301 | -36% | 0 | 0 | — |
case-06 | fail→pass | 13,223 | 5,615 | -58% | 1 | 1 | 0% | 2,486 | 1,572 | -37% | 0 | 0 | — |
case-07 | fail→pass | 9,995 | 2,605 | -74% | 1 | 1 | 0% | 1,824 | 1,028 | -44% | 0 | 0 | — |
case-08 | fail→fail | 3,672 | 2,189 | -40% | 1 | 1 | 0% | 628 | 920 | +46% | 0 | 0 | — |
case-09 | pass→pass | 6,711 | 1,385 | -79% | 1 | 1 | 0% | 999 | 794 | -21% | 0 | 0 | — |
case-10 | fail→fail | 7,429 | 1,571 | -79% | 1 | 1 | 0% | 1,184 | 812 | -31% | 0 | 0 | — |
case-11 | fail→pass | 5,047 | 1,219 | -76% | 1 | 1 | 0% | 850 | 746 | -12% | 0 | 0 | — |
case-12 | pass→pass | 3,822 | 2,347 | -39% | 1 | 1 | 0% | 825 | 1,043 | +26% | 0 | 0 | — |
case-13 | fail→pass | 11,072 | 3,108 | -72% | 1 | 1 | 0% | 1,800 | 1,116 | -38% | 0 | 0 | — |
case-14 | fail→fail | 13,005 | 7,406 | -43% | 1 | 1 | 0% | 2,810 | 851 | -70% | 0 | 0 | — |
case-15 | fail→fail | 15,958 | 4,054 | -75% | 1 | 1 | 0% | 3,255 | 1,200 | -63% | 0 | 0 | — |
case-16 | fail→fail | 13,352 | 4,175 | -69% | 1 | 1 | 0% | 2,799 | 891 | -68% | 0 | 0 | — |
case-17 | fail→pass | 13,217 | 9,225 | -30% | 1 | 1 | 0% | 2,548 | 2,007 | -21% | 0 | 0 | — |
case-18 | fail→fail | 16,557 | 14,142 | -15% | 1 | 1 | 0% | 3,753 | 3,377 | -10% | 0 | 0 | — |
case-19 | fail→fail | 14,567 | 23,910 | +64% | 1 | 1 | 0% | 2,731 | 4,185 | +53% | 0 | 0 | — |
case-20 | pass→pass | 14,880 | 8,030 | -46% | 1 | 1 | 0% | 2,684 | 2,308 | -14% | 0 | 0 | — |
case-21 | pass→pass | 11,577 | 7,828 | -32% | 1 | 1 | 0% | 2,048 | 2,193 | +7% | 0 | 0 | — |
case-22 | pass→pass | 11,294 | 8,919 | -21% | 1 | 1 | 0% | 2,092 | 2,333 | +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, and 19 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +27 percentage points is the difference between those two pass rates over the 19 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.