Consensus

Consensus enables a "model chorus" workflow that queries multiple AI providers in parallel, allowing for cross-validation and more reliable decision-making. By executing several models at once, users can choose from five synthesis strategies—such as majority voting, weighted detail, or structured comparison—to ensure that the final output is accurate and comprehensive. This skill is essential for high-stakes tasks, factual verification, and architectural reviews where a single model's perspective may be limited or prone to hallucinations, providing a robust framework for reliability through model agreement.

主な機能

ユースケース