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 agentmods add commands/4pixeltechbr/vibedev/vd-devils-advocategit clone --depth 1 https://github.com/4pixeltechBR/VibeDevWhat 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 | $0.00000 | $0.01581 |
| Opus 5 | $0.00000 | $0.00790 |
| Sonnet 5 | $0.00000 | $0.00316 |
| Haiku 4.5 | $0.00000 | $0.00158 |
Grade A, and why
vd-devils-advocate 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 3d 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 — 132 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/vd-devils-advocate
🇧🇷 PT-BR · 🇺🇸 EN follows
Sub-comando do
/vd-plan. Roda um pré-mortem na decisão ANTES de quebrar em sub-tarefas. Força o agente (e o usuário) a enxergar o que pode dar errado.
🇧🇷 O que faz
Ativado automaticamente como gate dentro do /vd-plan (opt-out via flag --skip-devils-advocate).
Lê o estado atual (PROJECT_STATE.md + ARCH_MAP.md se existir + a decisão que vai ser tomada) e gera 3 entregas obrigatórias:
- Top 3 riscos da decisão (probabilidade × impacto, com explicação)
- 1 contradição interna que o agente detectou entre o que o usuário quer e o que o estado do projeto diz
- 1 ângulo cego — uma coisa que ninguém (nem o agente, nem o usuário) mencionou, mas que pode implodir o plano
O agente NÃO pode prosseguir sem que o usuário responda explicitamente a cada item (aceitar, mitigar ou descartar).
🇺🇸 What it does
Auto-activated as a gate inside /vd-plan (opt-out via --skip-devils-advocate flag).
Reads current state (PROJECT_STATE.md + ARCH_MAP.md if exists + the decision about to be made) and produces 3 mandatory outputs:
- Top 3 risks of the decision (probability × impact, with explanation)
- 1 internal contradiction the agent detected between what the user wants and what project state says
- 1 blind angle — something nobody (not the agent, not the user) mentioned that could implode the plan
Agent CANNOT proceed without user explicitly responding to each item (accept, mitigate, or discard).
📋 Output format (append no PROJECT_STATE.md)
## Devils Advocate — AAAA-MM-DD HH:MM
<!-- Gerado por /vd-plan como gate. Não prosseguir sem resposta do usuário. -->
### Top 3 riscos
1. **[Risco]** — prob: [baixa|média|alta] · impacto: [baixo|médio|alto] · por quê: [explicação]
2. **[Risco]** — prob: [...] · impacto: [...] · por quê: [...]
3. **[Risco]** — prob: [...] · impacto: [...] · por quê: [...]
### Contradição interna detectada
- **Tensão:** [o que o usuário disse em X vs o que PROJECT_STATE diz em Y]
- **Resolução proposta:** [sugestão do agente]
### Ângulo cego
- **[Cenário que ninguém pensou]:** [descrição curta]
- **Por que pode ser foda:** [explicação]
### Resposta do usuário
<!-- Obrigatório para /vd-plan prosseguir. -->
- [ ] Risco 1: [aceito|mitigo|descarto] — [comentário]
- [ ] Risco 2: [aceito|mitigo|descarto] — [comentário]
- [ ] Risco 3: [aceito|mitigo|descarto] — [comentário]
- [ ] Contradição: [resolução escolhida]
- [ ] Ângulo cego: [incorporo|descarto] — [comentário]
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.
- 3d ago First seen · 132 lines · 0 tokens per session scan A c0ca82acb765
vd-devils-advocate is a command published in the GitHub repository 4pixeltechBR/VibeDev (16 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,581 tokens. 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
sync-linear
Sync current work with Linear ticket status.
django-model
Create Django models with proper structure and relationships.
constitution
通过交互式或提供的原则输入创建或更新项目宪法，确保所有依赖模板保持同步。.
release-notes
Generate release notes from git history — maintain a Keep a Changelog CHANGELOG.md entry and emit a user-facing release body, without deciding versions or mutating git state.
review
Karen repository review with market-aware scoring.
feature
End-to-end authoring workflow for The Claude Code Bible — verify claims against the real tool, split chapters into path-disjoint slices, draft with parallel agents in this ONE checkout (never worktrees), fix TOC + cross-links, commit by path and merge. Reads intent from the prompt.