Record structured thoughts with revision history for multi-model deliberation.
npx skills add https://github.com/Raudbjorn/cognitive-construct --skill rhetoric
Please help me install this Agent Skill. Skill: rhetoric Source: https://github.com/Raudbjorn/cognitive-construct/tree/main/rhetoric Command: npx skills add https://github.com/Raudbjorn/cognitive-construct --skill rhetoric
Complex problem-solving often leads to unstructured thinking, cognitive biases, and suboptimal decisions. This Skill provides a robust framework for structured reasoning, multi-perspective deliberation, and pattern analysis to overcome these challenges.
deliberate command to get a multi-faceted debate from different AI models, complete with a confidence-scored consensus.Deliberate on a critical question using multiple AI perspectives: python3 scripts/rhetoric.py deliberate "Should we use microservices or monolith?" --rounds 3
High-intent search queries and answers about installing and using this skill.
Structured reasoning for decisions involves recording thoughts with full revision history and orchestrating deliberation between multiple AI models to gain diverse perspectives. This Skill records branching thought paths, detects patterns that reveal cognitive biases, and generates consensus summaries—enabling you to trace how your reasoning evolved and identify blind spots before committing to major choices.
Yes. This Skill integrates VibeCheck pattern analysis to review your thinking patterns and surface common cognitive biases like analysis paralysis. By tracking revision history and branching thoughts across deliberation sessions, you identify recurring decision traps and interrupt them before they degrade outcome quality.
Use the deliberate command to orchestrate debates between multiple AI models on a single question, specify the number of rounds, and receive a confidence-scored consensus. The Skill manages session state, tracks each model's reasoning, generates transcripts, and produces summaries—streamlining collaborative problem-solving without manual coordination.
Deliberation produces automatic transcripts capturing all model contributions, structured summaries with confidence scores, full revision history showing how reasoning evolved, and pattern analysis identifying cognitive biases. These artifacts support audit trails, decision documentation, and downstream analysis of reasoning quality.
Yes. This Skill requires httpx for multi-model API orchestration and pydantic for structured thought and session data validation. Both are lightweight libraries; httpx handles concurrent calls to multiple AI endpoints, and pydantic ensures revision history and branching paths remain consistent throughout deliberation.
Use structured deliberation for high-stakes architectural or strategy decisions where cognitive bias, incomplete perspectives, or suboptimal reasoning poses significant risk. Multi-model debate is most valuable when the decision affects system design, resource allocation, or long-term commitments—cases where diverse reasoning and bias detection justify the deliberation overhead.
89jobrienOptimize LLM prompts with structured design, few-shot patterns, and chain-of-thought reasoning.
markus41Orchestrate multi-agent council reviews with configurable roles, weighted voting, and consensus detection.
Rune-kitPersist and recall decisions, patterns, errors, and insights in a semantic graph.
Smith-106Analyzes fictional characters' decisions using Munger's mental models and cognitive biases.