Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/georgemarmelstein/superjurista-devnpx agentmods add commands/georgemarmelstein/superjurista-dev/baixar-inteligenteWrote 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/commands/georgemarmelstein/superjurista-dev/baixar-inteligente)<a href="https://agentmods.dev/commands/georgemarmelstein/superjurista-dev/baixar-inteligente"><img src="https://agentmods.dev/badge/commands/georgemarmelstein/superjurista-dev/baixar-inteligente/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/commands/georgemarmelstein/superjurista-dev/baixar-inteligente"><img src="https://agentmods.dev/badge/commands/georgemarmelstein/superjurista-dev/baixar-inteligente.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 | $0.00013 | $0.01256 |
| Opus 5 | $0.00006 | $0.00628 |
| Sonnet 5 | $0.00003 | $0.00251 |
| Haiku 4.5 | $0.00001 | $0.00126 |
Grade A, and why
baixar-inteligente 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 9d 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 — 202 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/baixar-inteligente
Pipeline de download inteligente de processos do PJE com selecao por LLM.
Argumentos
$ARGUMENTS = <numero-processo> [--merge] [--converter]
Flags opcionais:
--merge Junta todos os PDFs em arquivo unico
--converter Converte para TXT (requer --merge)
Exemplos:
/baixar-inteligente 0822811-25.2019.4.05.8100
/baixar-inteligente 0822811-25.2019.4.05.8100 --merge
/baixar-inteligente 0822811-25.2019.4.05.8100 --merge --converter
EXECUCAO OBRIGATORIA
Etapa 0: Parse de argumentos
args = "$ARGUMENTS".split()
NUMERO_PROCESSO = args[0]
FLAG_MERGE = "--merge" in args
FLAG_CONVERTER = "--converter" in args
if FLAG_CONVERTER and not FLAG_MERGE:
FLAG_MERGE = True # Ativa automaticamente
WORKSPACE = f"data/sentenca/{NUMERO_PROCESSO}"
Criar diretorio:
mkdir -p data/sentenca/NUMERO_PROCESSO
Etapa 1: Verificar/Capturar sessao PJE (CONTEXTO ISOLADO)
Verificar se pje_session.json existe e e recente (< 8 horas).
Se invalido, usar skill capturar-sessao-pje que roda em contexto isolado:
Invocar skill: capturar-sessao-pje
A skill tenta Chrome MCP primeiro, e se falhar, pede HAR ao usuario.
Checkpoint: Arquivo pje_session.json existe?
Etapa 2: Identificar ID do processo
Executar EXATAMENTE:
python .claude/skills/pje-download/scripts/buscar_processo_por_numero.py --cookies pje_session.json --numero NUMERO_PROCESSO
Capturar o ID_PROCESSO retornado.
Etapa 3: Extrair indice completo
Executar EXATAMENTE:
python .claude/skills/pje-download/scripts/extrair_indice_completo.py --cookies pje_session.json --id-processo ID_PROCESSO --output WORKSPACE/indice_completo.json
Checkpoint: Arquivo contem lista de documentos?
Etapa 4: Selecao inteligente (Task tool)
Usar Task tool para spawnar agente seletor:
Task tool:
subagent_type: general-purpose
prompt: |
Leia o indice em WORKSPACE/indice_completo.json
Aplique a heuristica "mentalidade de juiz":
- NUCLEAR: Sempre incluir (peticao inicial, sentenca, laudo, embargos)
- SUBSTANTIVO: Analisar descricao (parecer, manifestacao)
- TEMPORAL: Incluir se recente (ultimos 6 meses)
- EXCLUIR: So com certeza absoluta
REGRA DE OURO: Na duvida, inclua.
Salve a selecao em WORKSPACE/selecao_documentos.json
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.
- 9d ago First seen · 202 lines · 13 tokens per session scan A 77db27bad156
baixar-inteligente is a command published in the GitHub repository georgemarmelstein/superjurista-dev (22 stars, last pushed 1mo ago), licensed MIT. It adds 13 tokens to every session and 1,256 once invoked, about $0.0001 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 commands, from other repositories
audit-export
Exportar audit trail en JSON/CSV/PDF para auditorías externas. ISO 42001 & NIST AI RMF compliant.
sign
Sign a PDF — stamp a signature block (name + date, or a PNG signature) on a signed copy.
family-tree
A tool that reads a Japanese family-register PDF and creates a self-contained HTML family relationship chart for inheritance matters. A family register is an official record used to establish family relationships and status.
finalize
A final submission workflow that protects a client's edits, creates a signed PDF, and prepares a filing package for the 02hearings location.
/a11y-report
Reporte de conformidad de accesibilidad para stakeholders y legal. Tres formatos: ejecutivo (resumen + score), técnico (detalles completos + código), legal (declaración VPAT/Section 508). Tracking de tendencias. Exportable.
audit
Generate a structured audit report for the current session or a specified time range. The report summarises every governed action, lists any blocked requests, and provides a compliance-ready summary you can download or share.