multi-persona-critique

To stress-test product strategy proposals, you can generate parallel critiques from five distinct philosophical personas that evaluate logical coherence, practicality, system complexity, user impact, and human agency risks, then synthesize the results into a consensus report.

Multi-agent critique for architecture decision reviews is a process where multiple independent personas evaluate proposals in parallel to avoid groupthink, providing per-proposal ratings and rankings to make tradeoffs explicit before synthesizing a consensus.

To format proposals for multi-persona evaluation, ensure all proposals are numbered and self-contained. This structured input allows each persona to independently rate, rank, and critique each idea without requiring external context.

Yes, you can validate feature brainstorming and detect groupthink by sending ideas to five distinct personas for independent evaluation. This process forces structured parallel critiques and preserves disagreements to highlight alignment and conflict.

The consensus synthesis report includes a consensus matrix with scored patterns, disagreement analysis, and final synthesis sections. It preserves both agreement and tension across the independent persona critiques to provide actionable guidance.

A limitation of using parallel critiques is that proposals must be strictly numbered and self-contained to function properly. Additionally, the process mandates per-proposal ratings and rankings for every persona, requiring highly structured input.