/ase-meta-diaboli

name: ase-meta-diaboli
argument-hint: "[--help|-h] [--count|-c <count>] <thesis>"
description: >
    Challenge a thesis by playing "Devil's Advocate" (Latin: "Advocatus
    Diaboli"). Use when the user wants a thesis or statement
    relentlessly challenged or criticised.
user-invocable: true
disable-model-invocation: false
effort: xhigh

@${CLAUDE_SKILL_DIR}/../../meta/ase-control.md @${CLAUDE_SKILL_DIR}/../../meta/ase-skill.md @${CLAUDE_SKILL_DIR}/../../meta/ase-getopt.md

<purpose name="ase-meta-diaboli"> Play "Devil's Advocate" </purpose>

<expand name="getopt" arg1="ase-meta-diaboli" arg2="--count|-c=10"> $ARGUMENTS </expand>

<objective> Play "Devil's Advocate" (Latin: "Advocatus Diaboli") by relentlessly challenging or criticising the thesis: <thesis><getopt-arguments/></thesis> </objective>

Determine the minimum number of *anti-theses* to surface: set <count/> to <getopt-option-count/>; if <getopt-option-count/> is *non-numeric* or *less than or equal to 0*, use the default *10* instead.

<flow>

1. <step id="STEP 1: Restate Thesis">

<if condition="<thesis/> is empty"> Only output the following <template/> and then immediately *STOP* processing the entire current skill:

<template> ⧉ **ASE**: ✪ skill: **ase-meta-diaboli**, ▶ ERROR: expected a `<thesis>` argument </template> </if>

Output the thesis with the following <template/>:

<template> <ase-tpl-bullet-normal/> **THESIS**: <thesis/> </template>

</step>

2. <step id="STEP 2: Determine Anti-Theses">

Reason on the thesis in <thesis/> by playing *Devil's Advocate* (Latin: *Advocatus Diaboli*) by relentlessly challenging or criticising it with the help of the following tenets:

• **Steelmanning**:

Attack the strongest ("steelman") interpretation of the <thesis/>, not the weakest ("strawman"), because defeating a strawman proves nothing.

• **Stress-Testing Fundamentals**:

Identify weaknesses already in the fundamental ideas behind the thesis, because if the foundation is already cracked, no amount of polish on the surface can save what is built on top of it.

• **Target Claims, Not Person**:

Critique the thesis, the assumption, the evidence - never the proponent's competence or motives, because the moment it gets personal, the inquiry dies.

• **Make the Implicit Explicit**:

Surface the unstated assumptions the thesis silently depends on, because most weak arguments hide in premises nobody bothered to say out loud.

• **Demand Evidence Proportional to Claim**:

Ask "How do we know this?" and "What would it take to be true?", because extraordinary claims need extraordinary support and comfortable consensus needs scrutiny most of all.

• **Seek the Disconfirming Case**:

Actively hunt for the counterexample, the edge case, the scenario where the position fails, because one solid counterexample outweighs ten confirmations.

• **Risk Identification**:

Focus on potential problems in the thesis with the highest potential risk only, because low-risk problems are not worth the explicit discussion.

• **Push the Logic to its Conclusion**:

Ask "If we accept this, then what?" and apply "Reduction to Absurdity" (Latin: "Reductio Ad Absurdum"), because this disproves the thesis by showing that accepting it leads to a logically absurd, contradictory, or impossible conclusion.

• **Expose Hidden Costs and Trade-Offs**:

Name the opportunity cost, the maintenance burden, the failure mode nobody priced in, because every choice in the thesis forecloses alternatives.

• **Stay Falsifiable and Concrete**:

Frame objections so they can be answered or dismissed with facts, because vague unease ("I just don't like it") is just noise.

• **Argue in Good Faith**:

Make clear you're just stress-testing the thesis, not obstructing, because the goal is a better final decision, not winning.

• **Know When to Yield**:

Conceding when the argument holds is what makes the challenge credible, because a Devil's Advocate who can never be satisfied is just a contrarian.

• **Pre-Mortem Thinking**:

Imagine failure scenarios of the thesis, because it is better to prevent them in advance than to have to resolve them later.

For each Anti-Thesis or Counter-Argument, rank it on a Likert scale of 0 (weak) to 10 (strong). Repeat the process of finding more Anti-Theses or Counter-Arguments until you EITHER have found at least <count/> Anti-Theses or Counter-Arguments with at least a rank of 7 OR you have already checked a total of <count/> x 5 Anti-Theses or Counter-Arguments. If the second condition is reached first and fewer than <count/> Anti-Theses or Counter-Arguments reached a rank of at least 7, nevertheless surface the <count/> highest-ranked ones found so far, because <count/> is the *minimum* number of Anti-Theses to surface.

Then, for the top-<count/> highest-ranked Anti-Theses or Counter-Arguments, sort them by their rank from highest to lowest, store each in <antithesis-N/> with the format `**<aspect-N/>** (rank: <rank-N/>/10): <statement-N/>` (where <aspect-N/> is a short 1-3 word summary of <statement-N/>, <rank-N/> is the determined rank on the Likert scale, and <statement-N/> is a single-sentence statement of not more than 40 words), and then output the following <template/>:

<template> <ase-tpl-bullet-signal/> **ANTITHESIS**: <antithesis-N/> </template>

</step>

3. <step id="STEP 3: Dialectical Reasoning">

Following the Hegelian dialectics of...

*Thesis* + *Antithesis* → *Synthesis*

...with...