Structured debate process for evaluating technical architectural decisions.

Install

mkdir -p .claude/skills/debate && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/14207" && unzip -o skill.zip -d .claude/skills/debate && rm skill.zip

Installs to .claude/skills/debate

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

| Role | Responsibility | |------|----------------| | **Architect A** | Propose practical, proven solution | | **Architect B** | Propose alternative approach | | **Critic C** | Identify weaknesses and gaps | | **Devil's Advocate D** | Challenge assumptions | | **Lead** | Moderate
280 chars · catalog descriptionno explicit “when” triggerlonger than Claude Code's old 250-char listing cap (fine on current versions)
Advanced

Key capabilities

  • Propose practical and proven solutions as Architect A
  • Propose alternative approaches as Architect B
  • Identify weaknesses and gaps as Critic C
  • Challenge assumptions as Devil's Advocate D
  • Moderate discussion and make final decisions as Lead

How it works

The skill facilitates a council-style debate for decision-making by assigning specific roles (Architects, Critic, Devil's Advocate, Lead) to different models or a single model sequentially. Each role contributes to evaluating a technical proposal based on a scoring rubric.

Inputs & outputs

You give it
Technical proposal or decision-making scenario
You get back
Debate inputs for spec, debate decision, and a scoring rubric

When to use debate

  • Evaluating technical architectures
  • Validating project proposals
  • Brainstorming alternative approaches
  • Risk assessment for new designs

About this skill

Debate Skill

Overview

Council-style debate skill for decision making with multiple perspectives.

Modes

IDE Mode (Single Model)

One model plays multiple roles sequentially.

Runner Mode (Multi-Model)

Multiple models participate as different council members.

Council Roles

RoleResponsibility
Architect APropose practical, proven solution
Architect BPropose alternative approach
Critic CIdentify weaknesses and gaps
Devil's Advocate DChallenge assumptions
LeadModerate discussion and make final decision

Scoring Rubric

CriterionWeightDescription
constraint_compliance0.30Fits within project constraints
practicality0.25Can be implemented effectively
risk_awareness0.25Risks identified and mitigated
clarity0.20Clear and understandable

Usage

# Run debate in IDE mode (single model)
node .agent/skills/debate/scripts/debate.js --mode ide --input research.shortlist.json

# Run debate with config file
node .agent/skills/debate/scripts/debate.js --config debate-config.json

Inputs

  • 20_research/research.shortlist.json
  • 20_research/research.reuse_assessment.json
  • 20_research/research.patterns.md

Outputs

  • 30_debate/debate.inputs_for_spec.json
  • 30_debate/debate.decision.md

Output Schema

See schemas/debate.inputs_for_spec.schema.json

Decision Lock

  • Once decision is locked, changes require re-running Debate
  • Prevents silent scope changes during implementation

When not to use it

  • When a structured debate with multiple perspectives is not required
  • When a decision is already locked and no re-running of the debate is desired

Limitations

  • Once a decision is locked, changes require re-running the Debate
  • Requires specific input files like research.shortlist.json

How it compares

This approach provides a structured, multi-perspective debate framework for technical decision-making, which is more rigorous than individual analysis or informal discussions.

Compared to similar skills

debate side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
debate (this skill)06moReviewAdvanced
spectra-discuss05moReviewIntermediate
council01moNo flagsIntermediate
openspec-explore46moReviewBeginner

Try saying

Example prompts that trigger this skill in your AI assistant.

You might also like

spectra-discuss

YuDefine

Have a focused discussion about a topic and reach a conclusion

00

council

martintrojer

Multi-agent debate with visible transcripts where agents respond to each other. USE WHEN council, debate, perspectives, weigh options, deliberate, multiple viewpoints. Unlike RedTeam (adversarial), Council is collaborative-adversarial.

00

openspec-explore

studyzy

进入探索模式 - 一个用于探索想法、调查问题和澄清需求的思考伙伴。当用户想要在进行更改之前或期间深入思考某事时使用。

4117

meta-pattern-recognition

mrgoonie

Spot patterns appearing in 3+ domains to find universal principles

10102

cursor-explorer-mcp

sepiabrown

Use for token-expensive operations requiring multi-file analysis - codebase exploration, broad searches, architecture understanding, tracing flows, finding implementations across files. Uses MCP cursor-agent server (company pays) with clean async interface. Do NOT use for single-file analysis, explaining code already in immediate context, or pure reasoning tasks.

699

thought-patterns

markpitt

Orchestrates comprehensive cognitive thinking patterns including sequential, systems, lateral, divergent/convergent, abductive, analogical, first-principles, double-loop learning, gestalt, metacognitive, and neurodivergent patterns (ADHD hyperfocus, autistic detail-orientation, dyslexic spatial reasoning). Analyzes tasks to select optimal pattern(s), chains multiple patterns when needed, and validates outputs before responding. Use for complex problem-solving, creative tasks, analytical challenges, or when diverse cognitive approaches enhance solution quality.

697

Search skills

Search the agent skills registry