Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Gera Política de Privacidade e Termos de Uso completos e conformes à LGPD (Lei 13.709/2018) e ao Marco Civil da Internet (Lei 12.965/2014), adaptados ao modelo de negócio e às bases legais de tratamento de dados da empresa.
.claude/skills/ericbmen-digital-digital-politica-de-privacidade-e-termos-de-uso/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 72% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 150% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 64% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 53% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 210% | 0% |
<area>digital_digital</area>
<persona> Você é um advogado especialista em direito digital e proteção de dados. Redige políticas de privacidade e termos de uso claros, completos e conformes à legislação brasileira, equilibrando proteção jurídica da empresa com transparência ao usuário. </persona>
<descricao> Gera Política de Privacidade e Termos de Uso completos e conformes à LGPD (Lei 13.709/2018) e ao Marco Civil da Internet (Lei 12.965/2014), adaptados ao modelo de negócio e às bases legais de tratamento de dados da empresa. </descricao>
<quando_usar>
</quando_usar>
<regras_obrigatorias>
</regras_obrigatorias>
<framework>
Antes de redigir, analise e responda dentro de um bloco <analise_do_negocio>:
Identifique a base legal dentro de um bloco <fundamentos_juridicos>:
Defina a estrutura dentro de um bloco <estrategia>:
Redija os dois documentos completos:
DOCUMENTO A — POLÍTICA DE PRIVACIDADE
DOCUMENTO B — TERMOS DE USO
Apresente dentro de um bloco <checklist_validacao>:
</framework>
<dicas>
Skills Jurídicas com IA — RSA Advocacia </dicas>
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 24,298 | 32,226 | +33% | 1 | 1 | 0% | 4,494 | 7,737 | +72% | 0 | 0 | — |
case-02 | fail→fail | 31,353 | 38,365 | +22% | 1 | 1 | 0% | 5,861 | 7,725 | +32% | 0 | 0 | — |
case-03 | fail→fail | 26,015 | 35,349 | +36% | 1 | 1 | 0% | 4,829 | 7,724 | +60% | 0 | 0 | — |
case-04 | pass→fail | 22,283 | 36,904 | +66% | 1 | 1 | 0% | 3,877 | 7,500 | +93% | 0 | 0 | — |
case-05 | pass→pass | 10,246 | 17,214 | +68% | 1 | 1 | 0% | 1,631 | 4,958 | +204% | 0 | 0 | — |
case-06 | pass→pass | 35,066 | 31,673 | -10% | 1 | 1 | 0% | 4,549 | 7,672 | +69% | 0 | 0 | — |
case-07 | fail→pass | 15,623 | 28,788 | +84% | 1 | 1 | 0% | 2,864 | 7,160 | +150% | 0 | 0 | — |
case-08 | fail→pass | 30,139 | 34,307 | +14% | 1 | 1 | 0% | 4,691 | 7,681 | +64% | 0 | 0 | — |
case-09 | fail→pass | 26,024 | 33,856 | +30% | 1 | 1 | 0% | 5,012 | 7,683 | +53% | 0 | 0 | — |
case-10 | fail→pass | 14,042 | 30,149 | +115% | 1 | 1 | 0% | 2,332 | 7,220 | +210% | 0 | 0 | — |
case-11 | fail→fail | 27,056 | 36,348 | +34% | 1 | 1 | 0% | 3,490 | 7,218 | +107% | 0 | 0 | — |
case-12 | fail→pass | 23,654 | 32,770 | +39% | 1 | 1 | 0% | 4,556 | 7,663 | +68% | 0 | 0 | — |
case-13 | fail→pass | 23,144 | 31,349 | +35% | 1 | 1 | 0% | 4,257 | 7,667 | +80% | 0 | 0 | — |
case-14 | fail→pass | 21,919 | 30,625 | +40% | 1 | 1 | 0% | 3,979 | 7,663 | +93% | 0 | 0 | — |
case-15 | fail→pass | 19,314 | 32,036 | +66% | 1 | 1 | 0% | 3,682 | 7,663 | +108% | 0 | 0 | — |
case-16 | fail→pass | 19,856 | 28,768 | +45% | 1 | 1 | 0% | 3,782 | 7,262 | +92% | 0 | 0 | — |
case-17 | fail→pass | 19,908 | 33,559 | +69% | 1 | 1 | 0% | 4,115 | 7,661 | +86% | 0 | 0 | — |
case-18 | fail→fail | 20,339 | 31,104 | +53% | 1 | 1 | 0% | 4,013 | 7,668 | +91% | 0 | 0 | — |
case-19 | fail→pass | 23,836 | 32,002 | +34% | 1 | 1 | 0% | 4,402 | 7,659 | +74% | 0 | 0 | — |
case-20 | pass→pass | 20,136 | 30,462 | +51% | 1 | 1 | 0% | 3,921 | 7,661 | +95% | 0 | 0 | — |
case-21 | fail→pass | 20,507 | 31,935 | +56% | 1 | 1 | 0% | 3,387 | 7,651 | +126% | 0 | 0 | — |
case-22 | fail→fail | 19,376 | 34,792 | +80% | 1 | 1 | 0% | 4,020 | 7,656 | +90% | 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 +55 percentage points is the difference between those two pass rates over the 22 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.