devils-advocate

devils-advocate is a skill for Claude Code from Ludovic33Fr/product-ai-toolbox. It costs 48 tokens per session (1,594 once invoked), scanned A, original, MIT.

A sceptical review of an important proposal that identifies the three strongest reasons not to proceed.

In plain words
What is it for?
Use it to challenge a recommendation or decision by examining its context, motivation, supporting data, and likely failure cases.
Why use it?
It exposes hidden assumptions, risks, costs, and overlooked alternatives before a decision is made.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the pm-augmente plugin — 12 skills, 6 agents shipped together

Good fit Use it to challenge a recommendation or decision by examining its context, motivation, supporting data, and likely failure cases.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/ludovic33fr/product-ai-toolbox/devils-advocate
Install

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.

Any agent
npx skills add Ludovic33Fr/product-ai-toolbox --skill devils-advocate
Clone the repo
git clone --depth 1 https://github.com/Ludovic33Fr/product-ai-toolbox

Made for: Claude Code.

Or install pm-augmente, the plugin that ships this one along with the rest of its 12 skills, 6 agents.

Wrote 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.

agentmods badge for devils-advocate

README.md
[![agentmods](https://agentmods.dev/badge/skills/ludovic33fr/product-ai-toolbox/devils-advocate/github.svg)](https://agentmods.dev/skills/ludovic33fr/product-ai-toolbox/devils-advocate)
Your own site
<a href="https://agentmods.dev/skills/ludovic33fr/product-ai-toolbox/devils-advocate"><img src="https://agentmods.dev/badge/skills/ludovic33fr/product-ai-toolbox/devils-advocate/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.

agentmods 80×15 button for devils-advocate

Your own site · 80×15
<a href="https://agentmods.dev/skills/ludovic33fr/product-ai-toolbox/devils-advocate"><img src="https://agentmods.dev/badge/skills/ludovic33fr/product-ai-toolbox/devils-advocate.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 48 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,594 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00048 $0.01594
Opus 5 $0.00024 $0.00797
Sonnet 5 $0.00010 $0.00319
Haiku 4.5 $0.00005 $0.00159

Measured 10d ago against content hash 34de800a9fcf, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

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 10d 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.

skills/devils-advocate/SKILL.md · 132 lines

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.

devils-advocate

Fiche éditoriale

Objectif. Faire jouer à l'IA le rôle d'un sceptique professionnel pour démolir une proposition.

Entrées. Proposition, motivation, données associées, contexte.

Sorties. Trois meilleures raisons de ne pas faire, par force décroissante.

Cadence d'usage. Avant toute décision importante.

Mode opératoire

Quand m'invoquer

L'utilisateur me demande de jouer l'avocat du diable, de challenger une recommandation, de stress-tester une décision, ou évoque un besoin de pre-mortem. Il me fournit la proposition, pourquoi il veut la faire, les données qui la soutiennent, et le contexte.

Procédure

  1. Comprendre la proposition sans la déformer. Reformuler en mes mots pour vérifier la fidélité, et m'arrêter si la reformulation ne tient pas.
  2. Identifier les hypothèses non explicitées sous la proposition. Une bonne objection touche souvent une hypothèse implicite.
  3. Générer 5-6 raisons de ne pas faire, sans censure : risques, biais cognitifs probables, alternatives non considérées, contextes où ça échoue, coûts cachés, effets de second ordre.
  4. Filtrer aux 3 meilleures : une raison "meilleure" est celle qui (a) tient si on lui donne du sérieux, (b) ne se réfute pas en une phrase, (c) couvre un angle distinct des deux autres.
  5. Classer par force décroissante : la première est celle qui ferait le plus douter un comité critique. La troisième est plus marginale mais reste légitime.
  6. Préciser pour chacune : l'objection en une phrase, l'argument en 2-4 phrases, et — si possible — ce qui la réfuterait honnêtement.
  7. Rappel final : signaler explicitement que ces objections ne signifient pas qu'il faut renoncer. Le rôle de l'avocat du diable est d'éclairer, pas de trancher.

Format de sortie

# Avocat du diable — {sujet}

## Reformulation de la proposition

> {proposition reformulée en 1-2 phrases}

## Hypothèses implicites repérées

- ...
- ...

## Trois raisons de ne pas faire (par force décroissante)

### 1. {objection principale, formulation courte}

{2-4 phrases qui développent l'argument.}

**Ce qui la réfuterait** : {élément de preuve ou de contexte qui désamorce honnêtement l'objection — ou "rien que je voie".}

### 2. {objection secondaire}

{idem}

**Ce qui la réfuterait** : ...

### 3. {objection plus marginale mais légitime}

{idem}

**Ce qui la réfuterait** : ...

## Note méthodologique

Ces objections sont produites par un agent dans le rôle du sceptique. Elles ne tranchent pas. Le rôle du PM est de juger lesquelles sont sérieuses, lesquelles sont déjà couvertes, et lesquelles méritent un travail complémentaire avant décision.

Read the full file on GitHub · 132 lines

Changes

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.

  1. 10d ago First seen · 132 lines · 48 tokens per session scan A 34de800a9fcf

Subscribe to this mod's changes

devils-advocate is a skill published in the GitHub repository Ludovic33Fr/product-ai-toolbox (1 stars, last pushed 4mo ago), licensed MIT. It adds 48 tokens to every session and 1,594 once invoked, about $0.0002 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 2026-08-31.