Run three-agent debates to produce a binding consensus document.
npx skills add https://github.com/dirien/yet-another-agent-harness --skill decide-dirien
Please help me install this Agent Skill. Skill: decide Source: https://github.com/dirien/yet-another-agent-harness/tree/main/.claude/skills/decide Command: npx skills add https://github.com/dirien/yet-another-agent-harness --skill decide-dirien
This Skill provides a structured, tri-agent debate framework to surface arguments, identify gaps, and reach a defensible binding decision.
Provide a decision topic and supporting context, then run Phase 1 to generate position papers.
High-intent search queries and answers about installing and using this skill.
Structured dialogue for architecture decisions uses a tri-agent debate framework with adversarial steelman arguments to surface risks and synthesize a binding consensus document.
Risk assessment debates for product strategy run through phase-driven workflows where two advocates present opposing positions and a moderator consolidates the final consensus.
Multi-agent debate resolves policy design tradeoffs by applying iterative convergence steps, where a moderator identifies gaps and questions to reach a defensible decision.
Start tri-agent tradeoff analysis by providing a decision topic and supporting context to trigger Phase 1, which generates initial position papers for the debate.
Adversarial debate frameworks are needed for decision-making scenarios requiring structured argumentation and risk mitigation, such as evaluating high-stakes product strategy or architecture choices.
Moderator synthesis for consensus building is limited to decision topics requiring structured argumentation, meaning it may not suit unstructured problems lacking clear opposing positions for iterative convergence.
wjgoarxivRun adversarial two-researcher debates with judge-scored rounds and structured reports.
nobelkConvert Product Requirements Documents into RFC-style Developer Design Documents.
hpsgdIdentify weaknesses in ideas or plans through adversarial analysis.
EvoClawCoordinate iterative multi-agent discussions to converge on a resolved artifact.
