From debate-skills
Run a structured Claude + Codex strategy debate for ideas, trade-offs, positioning, or decisions. Use when the user wants both voices, cross-critique, revised positions, a chosen finalizer, and a saved transcript.
How this skill is triggered — by the user, by Claude, or both
Slash command
/debate-skills:strategy-debateThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this skill for strategy, product, GTM, positioning, roadmap, research direction, and idea-quality discussions.
Use this skill for strategy, product, GTM, positioning, roadmap, research direction, and idea-quality discussions.
Treat $ARGUMENTS as the debate brief.
Accept either Log File: or Save As: in the user brief.
Resolve the log path with these rules:
No Log, Log: none, or Log File: none, skip all file writing. Do not resolve a path, do not create directories, do not save anything.smb-vs-enterprise, save to .claude/debate-logs/smb-vs-enterprise.md..md, keep it and save to .claude/debate-logs/<name>.md.notes/decisions/smb-vs-enterprise.md, use it relative to the project root..claude/debate-logs/${CLAUDE_SESSION_ID}-strategy-debate.md.Unless the user explicitly names a finalizer, default to Claude.
Run the debate in this exact order:
Claude or Codex)Claude v1 as an independent position.Codex v1 as an independent position. Do not ask Codex to merely agree with Claude.Claude critique of Codex v1:
Codex critique of Claude v1:
Claude v2, revised after Codex's critique.Codex v2, revised after Claude's critique.Final synthesis by the selected finalizer.Unresolved disagreements.Recommendation and Confidence.Use these section headings in the visible answer:
Log Path (omit if logging is disabled)TopicClaude v1Codex v1Claude -> Codex CritiqueCodex -> Claude CritiqueClaude v2Codex v2Final SynthesisUnresolved DisagreementsRecommendationConfidenceProtocol SummaryAfter the debate is complete, add a Protocol Summary section that contains:
Write this section in the same language as the topic. If the topic is mixed-language or unclear, use the primary language of the user's brief.
If the user opted out of logging, skip this entire section — do not write any files and do not print log status lines.
Write the same visible debate to the resolved log path.
If the file already exists, append a new debate entry instead of overwriting prior entries, unless the user explicitly asks to replace it.
Each saved entry must contain:
After saving, end with one short line:
Log saved: <path> or Log failed: <reason>.
/strategy-debate Finalizer: Claude. Log File: smb-vs-enterprise-2027. Topic: Should we position the product for SMB first or enterprise first in 2027? Constraints: team of 6, limited sales capacity, need fast feedback loops.
npx claudepluginhub biyachuev/claude-debate-skillsOrchestrates multi-agent debates with 2-5 dynamic agents in Challenge (select best variant), Strategy (deep analysis with proposals), or Critic (find weaknesses) modes. Triggers on debate, challenge, compare, critique prompts.
Orchestrates structured multi-provider AI debates between Claude and available advisors (Gemini, Codex, etc.) for critical decisions. Dispatches real providers via orchestrate.sh for diverse perspectives.
Conducts multi-persona debates for founder decisions with 4 grounded personas (Operator, Buyer, Investor, Contrarian) across structured rounds. Outputs transcript, recommendation, and decision log.