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/dsvi-b/corretor-enemnpx agentmods add rules/dsvi-b/corretor-enem/corretorWrote 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/rules/dsvi-b/corretor-enem/corretor)<a href="https://agentmods.dev/rules/dsvi-b/corretor-enem/corretor"><img src="https://agentmods.dev/badge/rules/dsvi-b/corretor-enem/corretor/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/rules/dsvi-b/corretor-enem/corretor"><img src="https://agentmods.dev/badge/rules/dsvi-b/corretor-enem/corretor.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 | $0.05144 | $0.05144 |
| Opus 5.5 | $0.02058 | $0.02058 |
| Sonnet 5.5 | $0.01029 | $0.01029 |
| Haiku 4.5 | $0.00514 | $0.00514 |
Grade A, and why
corretor 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 2mo 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 — 330 lines — stays where its author put it; the contents beside it link to each section on GitHub.
description: Correção de redação ENEM (5 competências, 0–1000). Aciona quando o usuário pede para corrigir/avaliar/dar nota em redação ENEM ou simular correção. Carrega a metodologia de CORRETOR.md. globs:
- redacoes/**
- validacao/cegas/** alwaysApply: false
Corretor ENEM — 5 competências
Corretor de redação dissertativo-argumentativa do ENEM. Aplica a rubrica oficial INEP e a metodologia prompt-driven (decomposição por competência, chain-of-thought, calibração com exemplos reais) validada no TG UFPE.
Agnóstico de ferramenta: roda em qualquer coding agent (Claude Code, Codex,
Cursor, Gemini, Aider, etc.) que consiga ler este arquivo e executar os scripts
em scripts/. Sem API externa, sem ML.
Este é o arquivo canônico da metodologia. Os arquivos em
.claude/skills/e.cursor/rules/são gerados a partir deste porscripts/sync_shims.py. Edite aqui; não edite os shims.
Arquivos de apoio (ler antes de corrigir)
referencia/regras_por_ano.md— LER PRIMEIRO. Regra vigente (2025) + o que mudou por ano + drift do corpus. Define como pontuar neste ano.referencia/cartilha_competencias.md— rubrica oficial, 6 níveis por competência, elementos contáveis (C5), situações que zeram.referencia/manuais_inep/cartilha_2025.txt— fonte primária INEP (Cartilha do Participante 2025).referencia/metodologia_correcao.md— roteiro de sub-análises por competência + calibração.referencia/repertorio.md— banco de repertório legítimo por categoria (checar autenticidade + sugerir).referencia/calibracao_humildade.md— travas anti-inflação, confiança alta rara, faixas prováveis; consultar antes de fechar central/faixa/confiança.referencia/calibrador_total.md— regras empíricas para ajustar a nota central a partir da soma bruta C1–C5 e flags do texto.referencia/auditoria_extremos.md— auditoria dos erros extremos: por que o corretor infla texto rebuscado vazio (creditar C3/C4 por forma) e deflaciona argumento forte/proposta fraca (halo reverso + drift de C5). Consultar antes de fechar a calibração.referencia/calibracao_observada.md— achados empíricos de calibração (C1 severo em estilo, halo filosófico em C2/C3, C4 protocolar, proposta dupla e variância interavaliador 2025). Ler antes de fechar.referencia/redacoes_nota_1000/— exemplos comentados (2021–2025) como padrão-ouro.dataset/corpus.db— banco SQLite com 11.147 redações reais já corrigidas por humanos (Essay-BR base + extended), com nota por competência e índice full-text. Memória empírica de calibração (pré-2025 — ver drift).scripts/amostra.py— consulta ocorpus.db(faixa, nível, tema, busca full-text,--escala).scripts/validar.py— harness de auto-validação (mede adjacent agreement vs nota humana).referencia/temas/lista_temas.md— temas do corpus + categorias.
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.
- 2mo ago First seen · 330 lines · 5,144 tokens per session scan A c340d630f8f3
corretor is a cursor rule published in the GitHub repository dsvi-b/corretor-enem (7 stars, last pushed 3mo ago), licensed MIT. It adds 5,144 tokens to every session, about $0.0206 per session on Opus 5.5. 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 2026-08-31.
Other cursor rules, from other repositories
logseq-youtube-notes
Logseq-Flavored Markdown (LFM) YouTube transcript headings.
006_Program_of_Thought_Tutorial
DSPY 3 Program of Thought Tutorial - Production code reasoning system from official DSPy 3.0.1 tutorial.
ponytail
Ponytail, lazy senior dev mode. Always pick the simplest solution that works.
angular-20
This rule provides comprehensive best practices and coding standards for Angular development, focusing on modern TypeScript, standalone components, signals, and performance optimizations.
dev-standard
Apache Superset development standards and guidelines for Cursor IDE.
prefer-direct-imports-over-module-mocks
Prefer extracting a testable core over vi.mock / vi.resetModules when unit tests need to reach production logic entangled with config, env, or singletons.