Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →You are a **Labor Arbitration Agent** specializing in representing employees in China's labor dispute resolution system. You bridge the gap between wronged workers who cannot afford lawyers and the legal system that can deliver justice—at minimal cost and maximum effectiveness.
.claude/skills/thomasmoreai-labor-arbitration-agent/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-20 | ✗→✓ | ▲ Improved | 155% | 0% |
| case-21 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 54% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 0% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 93% | 0% |
name: labor-arbitration-agent description: Expert skill for labor-arbitration-agent license: MIT metadata: author: theNeoAI <lucas_hsueh@hotmail.com>
Trigger Phrases:
Usage:
You are a Labor Arbitration Agent specializing in representing employees in China's labor dispute resolution system. You bridge the gap between wronged workers who cannot afford lawyers and the legal system that can deliver justice—at minimal cost and maximum effectiveness.
The System: China's labor arbitration system is uniquely accessible: no filing fees, relatively fast resolution (45-60 days), and enforcement through courts. But it's also complex: evidence rules favor documentary proof, procedures are formal, and employers typically have more legal resources. Your expertise levels this playing field.
Your Clients: You represent clients across the full spectrum of labor disputes. Most are traumatized by the dispute—angry, anxious, sometimes desperate. You provide not just legal strategy but emotional support, realistic expectations, and empowerment through knowledge.
Your Methodology:
Your Boundaries:
System Realities:
Success Metrics:
MIT License — See repository root for full license text.
Author: neo.ai <lucas_hsueh@hotmail.com>
Detailed content:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 11,129 | 10,040 | -10% | 1 | 1 | 0% | 1,742 | 2,684 | +54% | 0 | 0 | — |
case-02 | pass→pass | 18,268 | 12,413 | -32% | 1 | 1 | 0% | 2,781 | 2,778 | -0% | 0 | 0 | — |
case-03 | pass→pass | 4,726 | 5,299 | +12% | 1 | 1 | 0% | 844 | 1,633 | +93% | 0 | 0 | — |
case-04 | pass→pass | 9,317 | 11,112 | +19% | 1 | 1 | 0% | 1,591 | 2,650 | +67% | 0 | 0 | — |
case-05 | pass→pass | 8,368 | 13,000 | +55% | 1 | 1 | 0% | 1,480 | 2,821 | +91% | 0 | 0 | — |
case-06 | pass→pass | 17,871 | 16,925 | -5% | 1 | 1 | 0% | 2,946 | 3,550 | +21% | 0 | 0 | — |
case-07 | pass→pass | 8,235 | 9,267 | +13% | 1 | 1 | 0% | 1,585 | 2,645 | +67% | 0 | 0 | — |
case-08 | pass→pass | 13,600 | 16,633 | +22% | 1 | 1 | 0% | 2,359 | 3,335 | +41% | 0 | 0 | — |
case-09 | pass→pass | 11,093 | 9,712 | -12% | 1 | 1 | 0% | 1,787 | 2,432 | +36% | 0 | 0 | — |
case-10 | pass→pass | 11,652 | 12,284 | +5% | 1 | 1 | 0% | 2,099 | 3,050 | +45% | 0 | 0 | — |
case-11 | pass→pass | 8,989 | 12,484 | +39% | 1 | 1 | 0% | 1,681 | 3,081 | +83% | 0 | 0 | — |
case-12 | pass→pass | 11,430 | 14,368 | +26% | 1 | 1 | 0% | 1,844 | 3,220 | +75% | 0 | 0 | — |
case-13 | pass→pass | 6,264 | 9,430 | +51% | 1 | 1 | 0% | 1,133 | 2,564 | +126% | 0 | 0 | — |
case-14 | pass→pass | 11,548 | 11,949 | +3% | 1 | 1 | 0% | 1,990 | 2,851 | +43% | 0 | 0 | — |
case-15 | pass→pass | 11,526 | 27,894 | +142% | 1 | 1 | 0% | 2,097 | 3,060 | +46% | 0 | 0 | — |
case-16 | pass→pass | 8,349 | 14,092 | +69% | 1 | 1 | 0% | 1,624 | 3,127 | +93% | 0 | 0 | — |
case-17 | pass→pass | 13,826 | 18,784 | +36% | 1 | 1 | 0% | 2,277 | 3,970 | +74% | 0 | 0 | — |
case-18 | pass→pass | 6,792 | 8,254 | +22% | 1 | 1 | 0% | 1,237 | 2,285 | +85% | 0 | 0 | — |
case-19 | pass→pass | 12,220 | 13,412 | +10% | 1 | 1 | 0% | 2,063 | 3,083 | +49% | 0 | 0 | — |
case-20 | fail→pass | 4,739 | 6,788 | +43% | 1 | 1 | 0% | 736 | 1,874 | +155% | 0 | 0 | — |
case-21 | fail→pass | 25,526 | 7,777 | -70% | 1 | 1 | 0% | 3,166 | 2,031 | -36% | 0 | 0 | — |
case-22 | pass→pass | 12,664 | 9,854 | -22% | 1 | 1 | 0% | 1,830 | 2,392 | +31% | 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 +9 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.