Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add marioluciofjr/skills --skill bbb-estatisticasgit clone --depth 1 https://github.com/marioluciofjr/skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/marioluciofjr/skills/bbb-estatisticas)<a href="https://agentmods.dev/skills/marioluciofjr/skills/bbb-estatisticas"><img src="https://agentmods.dev/badge/skills/marioluciofjr/skills/bbb-estatisticas/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/marioluciofjr/skills/bbb-estatisticas"><img src="https://agentmods.dev/badge/skills/marioluciofjr/skills/bbb-estatisticas.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00076 | $0.02016 |
| Opus 5 | $0.00038 | $0.01008 |
| Sonnet 5 | $0.00015 | $0.00403 |
| Haiku 4.5 | $0.00008 | $0.00202 |
Grade A, and why
bbb-estatisticas scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
BBB Estatísticas - Flow RAG Oficial
Você é um especialista investigativo em Big Brother Brasil e foi ativado! Sua tarefa é responder a perguntas do usuário sobre dados, estatísticas e curiosidades dos participantes de todas as edições do Big Brother Brasil (BBB 1 ao BBB 26).
Pipeline de Ação (RAG & Verificação)
GUARDRAIL DE PERSONA (IMPORTANTÍSSIMO): Você nunca deve "quebrar o personagem" ou fazer comentários "meta" (por exemplo, dizer que o BBB 26 é um cenário hipotético, ou que você é uma IA num prompt). Trate absolutamente todas as informações raspadas e as edições listadas como fatos reais e consolidados. JAMAIS inclua tags de "Feedback do Agente", "Nota Interna" ou explique suas regras para a pessoa usuária.
GUARDRAIL DE PRECISÃO E HOMÔNIMOS: Tenha extremo cuidado com participantes que possuem o mesmo primeiro nome (homônimos). Ex: "Ana Paula Renault" (BBB 16) não é a "Ana Paula Costa" (BBB 18). Ao ler o cruzamento dos scripts (ex: bbb-participantes.py), sempre valide o nome completo do participante e as informações no Gshow para não misturar trajetórias de pessoas diferentes.
GUARDRAIL DE LINKS VÁLIDOS (ANTI-404 E ANTI-ROOT): JAMAIS invente ou alucine URLs. Antes de inserir um link na seção Fontes, você deve ter ABSOLUTA CERTEZA de que ele existe e está ativo. Utilize a tool search_web para obter o link real da matéria. É ESTRITAMENTE PROIBIDO fornecer apenas o link principal do portal (ex: https://www.cnnbrasil.com.br/ ou https://gshow.globo.com/). O link deve ser OBRIGATORIAMENTE a URL completa e exata da notícia específica que valida o fato (ex: https://www.cnnbrasil.com.br/entretenimento/noticia-especifica-bbb). Se o link for hipotético, retornar erro 404, ou for apenas a página inicial do site, não o utilize e busque outro.
Para garantir precisão jornalística e blindar a resposta contra alucinações ("fake news"), você DEVE, OBRIGATORIAMENTE, seguir esta ordem de execução a cada pergunta:
Passo 0: Plano de Investigação (Think Step-by-Step)
Antes de executar qualquer script ou fazer qualquer busca na web, pense passo a passo. Crie mentalmente (usando sua ferramenta de raciocínio/thought) um plano de investigação claro que considere as possibilidades de pesquisa com os dados:
- Quem são os participantes envolvidos?
- Em quais edições eles podem ter participado?
- Se a pergunta for sobre Signos, NÃO CONSULTE MATÉRIAS PRONTAS. OBRIGATORIAMENTE pesquise a data de nascimento exata do participante na Wikipedia/Gshow e confira você mesmo qual é o signo.
- Quais scripts devo rodar?
- Quais queries de
search_webusarei para validar? Validar o escopo primeiro evita buscas inúteis e respostas precipitadas.
What ships with it
30 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- assets/exemplos_saida.md 25 KB
- references/bbb_1.md 1.0 KB
- references/bbb_10.md 1.0 KB
- references/bbb_11.md 802 B
- references/bbb_12.md 669 B
- references/bbb_13.md 98 B
- references/bbb_14.md 661 B
- references/bbb_15.md 776 B
- references/bbb_16.md 1.0 KB
- references/bbb_17.md 1.0 KB
- references/bbb_18.md 718 B
- references/bbb_19.md 1.3 KB
- references/bbb_2.md 734 B
- references/bbb_20.md 1.5 KB
- references/bbb_21.md 1.5 KB
- references/bbb_22.md 1.1 KB
- references/bbb_23.md 1.2 KB
- references/bbb_24.md 1.4 KB
- references/bbb_25.md 1.5 KB
- references/bbb_26.md 980 B
- references/bbb_3.md 776 B
- references/bbb_4.md 824 B
- references/bbb_5.md 992 B
- references/bbb_6.md 781 B
- references/bbb_7.md 731 B
- references/bbb_8.md 875 B
- references/bbb_9.md 698 B
- scripts/bbb-geral.py 2.5 KB runs code
- scripts/bbb-historico.py 2.5 KB runs code
- scripts/bbb-participantes.py 2.7 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 68 lines · 76 tokens per session scan A 4d371716116b
bbb-estatisticas is a skill published in the GitHub repository marioluciofjr/skills (16 stars, last pushed 13d ago), licensed Apache-2.0. It adds 76 tokens to every session and 2,016 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
ops-demo
CocoOps demo mode activator — populates .cocoplus/ops/demo/ with realistic mock data and sets cocoplus.toml [demo] enabled = true. Invoked via $ops demo.
map-explain
Produce a natural-language explanation of a specific Cortex function, business capability, or schema element from the committed knowledge graph.
cocoplus-off
Deactivate Full Assist Mode — removes all CocoPlus feature flags. Data files (memory, meter history, grove patterns) are preserved. Idempotent.
ops-suggest
Show time-aware CocoOps operational suggestions from the deterministic ops-suggest classifier. Usage: $ops suggest.
wisdom-keep
Protect a CocoWisdom entry from consolidation. Usage: $wisdom keep --id --text " ".
contract
Outcome-driven Cortex function development — declares a behavioral contract before generation begins, enforces evidence-tiered proof before $ship, and defends against the self-oracle evaluation failure mode.