agentleFS
stepaniniwisstepaniniwis/claude-skills/skills/build/argument-architecture/SKILL.md

argument-architecture

Skill0 starsChanged 7 months agoMITView on GitHub

Strategic positioning and argumentation for academic papers. Combines Swales' CARS model (macro-level paper architecture, Introduction design, research space creation) with Toulmin's argumentation model (micro-level claim-evidence-warrant structure). Use when developing theoretical contributions, sharpening the 'So what?' question, structuring Introductions, building persuasive arguments, or diagnosing why an argument feels weak. Triggers on requests involving contribution, positioning, gap identification, argument structure, or theoretical framing.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create the space your paper fills.

    line 38
  2. Distinctive:

    Every academic claim needs:

    line 78
  3. Distinctive:

    Not all gaps are equal. Strong contributions:

    line 128
  4. Distinctive:

    When writing academic texts, apply skills hierarchically:

    line 161

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Argument Architecture Skill

Strategic positioning (CARS) + Argumentative persuasion (Toulmin) = Publishable scholarship.

Two-Level Architecture

LevelModelFunctionUnit
MacroCARSPosition paper in scholarly conversationSections, especially Introduction
MicroToulminBuild persuasive individual argumentsSentences, paragraphs

CARS answers: "Why should anyone read this paper?"
Toulmin answers: "Why should anyone believe this claim?"

Swales' CARS Model (Macro)

Create A Research Space — The Introduction as strategic argument.

Move 1: Establish Territory

Convince readers the topic matters.

StepFunctionSignal Phrases
1aClaim centrality"X is a fundamental concern...", "X plays a crucial role..."
1bTopic generalizations"X typically involves...", "Research on X has shown..."
1cReview prior research"Previous work establishes...", "Scholars have demonstrated..."

Purpose: Reader concludes "Yes, this is important territory."

Move 2: Establish Niche (The Gap)

Create the space your paper fills. This is the critical move.

StepFunctionSignal Phrases
2aCounter-claim"However, this approach fails to...", "Yet these accounts overlook..."
2bIndicate gap"What remains unclear is...", "Less attention has been paid to..."
2cRaise questions"This raises the question...", "It remains to be determined..."
2dContinue tradition"Following X's call for...", "Extending this line of inquiry..."

Purpose: Reader concludes "Something is missing here."

Move 3: Occupy Niche (Your Contribution)

Present what you offer.

StepFunctionSignal Phrases
3aOutline purpose"This paper argues that...", "We propose..."
3bAnnounce findings"The analysis reveals...", "We demonstrate..."
3cIndicate structure"The paper proceeds as follows...", "Section 2 develops..."

Purpose: Reader concludes "This paper fills the gap."

For extended guidance: references/cars-model.md

Toulmin Model (Micro)

Six components of every argument:

ComponentQuestionRole
ClaimWhat are you asserting?The conclusion you want accepted
DataWhat's your evidence?Facts, citations, observations
WarrantWhy does data support claim?The logical bridge (often implicit)
BackingWhat supports the warrant?Theory, methodology, authority
QualifierHow certain?Scope conditions, hedges
RebuttalWhat are the limits?Exceptions, counterarguments

Minimum Viable Argument

Every academic claim needs:

  1. Claim — explicit assertion
  2. Data — cited evidence
  3. Warrant — defensible connection (can be implicit if obvious)

Argument Diagnosis

SymptomDiagnosisTreatment
"So what?"Missing warrantExplain why evidence matters
"Says who?"Missing dataAdd citations
"Always?"Missing qualifierAdd scope conditions
"But what about...?"Missing rebuttalAddress objection
"Why should I believe that?"Missing backingGround the warrant

For extended guidance: references/toulmin-model.md

Integration: CARS → Toulmin

Each CARS move contains Toulmin arguments:

Move 1 as Argument

  • Claim: This topic matters
  • Data: Prior research, real-world significance
  • Warrant: Scholarly attention indicates importance

Move 2 as Argument

  • Claim: A gap exists
  • Data: What's missing in existing work
  • Warrant: This gap matters because...
  • Rebuttal: Acknowledge what existing work does accomplish

Move 3 as Argument

  • Claim: This paper fills the gap
  • Data: Preview of findings/framework
  • Warrant: Our approach succeeds where others failed
  • Qualifier: Scope and limitations

The "So What?" Test

Apply at three levels:

LevelQuestionIf Weak
PaperWhy does this contribution matter?Sharpen Move 2 gap
SectionWhy is this section here?Connect to thesis
ParagraphWhy does this claim matter?Add warrant

Contribution Types

Not all gaps are equal. Strong contributions:

TypeWhat You DoSignal
Theoretical extensionAdd to existing framework"We extend X by..."
Theoretical synthesisIntegrate separate streams"Bringing together X and Y reveals..."
Boundary conditionsShow where theory breaks"X holds only when..."
Conceptual clarificationResolve ambiguity"We distinguish between..."
New mechanismExplain how/why"The mechanism connecting X to Y..."
ProblematizationChallenge assumptions"Contrary to received wisdom..."

Quick Diagnostics

Introduction Weak?

  • Move 1 present? (Territory established?)
  • Move 2 sharp? (Gap clear and consequential?)
  • Move 3 responsive? (Contribution addresses the gap?)
  • Moves in sequence? (Territory → Gap → Contribution?)

Argument Weak?

  • Claim explicit?
  • Data present and cited?
  • Warrant defensible?
  • Qualifier appropriate?
  • Obvious rebuttals addressed?

"So What?" Failing?

  • Gap is trivial → Find a more consequential gap
  • Warrant missing → Explain why evidence matters
  • No stakes → Connect to broader concerns

Skill Hierarchy: When to Use What

When writing academic texts, apply skills hierarchically:

┌─────────────────────────────────────────────────────────┐
│  CARS (Macro)                                           │
│  → Governs SECTION logic                                │
│  → Positions paper in literature                        │
│  → Structures Introduction, frames entire argument      │
├─────────────────────────────────────────────────────────┤
│  TOULMIN (Meso)                                         │
│  → Governs ARGUMENT logic                               │
│  → Structures claims within Theory & Discussion         │
│  → Ensures each argument is complete and defensible     │
├─────────────────────────────────────────────────────────┤
│  ACADEMIC-WRITING / PEEL (Micro)                        │
│  → Governs PHRASING                                     │
│  → Realizes already-defined arguments in prose          │
│  → Handles style, citation integration, paragraph flow  │
└─────────────────────────────────────────────────────────┘

Decision Rules

QuestionSkill
"Where does this section fit in the paper?"CARS
"Is this argument complete and persuasive?"Toulmin
"Is this paragraph well-written?"Academic-Writing

Workflow Sequence

  1. Plan with CARS: What territory? What gap? What contribution?
  2. Structure with Toulmin: What claims? What data? What warrants?
  3. Write with PEEL: Point, Evidence, Explanation, Link

Critical: PEEL realizes arguments—it does not invent or justify them. If a paragraph feels weak, diagnose with Toulmin first (missing warrant? missing data?), then fix the prose with PEEL.

Integration Example

CARS decides: "Section 3 occupies the niche by presenting my theoretical framework."

Toulmin structures: "The section's main claim is X, supported by data Y, warranted by Z, qualified by W."

PEEL executes: "Paragraph 1 states the claim (Point), presents evidence (Evidence), explains significance (Explanation), transitions to next argument (Link)."

Reference Files

  • references/cars-model.md: Extended CARS with examples
  • references/toulmin-model.md: Toulmin patterns and diagnostics
  • references/contribution-types.md: Theoretical contribution patterns

Mandatory Quality Layer: Anti-AI Writing

This skill always activates the anti-ai-writing skill as a non-negotiable filter.

Argument architecture and AI-pattern avoidance are not separate concerns — they are the same concern at different levels of resolution. The standard AI tells (present-participle inflation, comprehensiveness-by-triplet, transition-word synthesis) are structural symptoms of arguments that have not fully committed to a position. When fixing surface-level AI patterns, check whether the underlying argumentative move is also incomplete.

Apply standard prohibitions + Academic Module on all outputs. Pay specific attention to:

  • Gap statements that describe absence vaguely ("little attention has been paid to...") rather than specifying what is missing and why it matters
  • Contribution statements hedged into near-invisibility ("this paper aims to contribute to...")
  • CARS moves that acknowledge existing work without taking a clear position on it

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.