directors

A multi-perspective review coordinates five expert directors to evaluate a project in parallel, synthesizing consensus, conflicts, and recommended actions for architecture or strategy decisions. You provide a project description and receive a synthesized board verdict.

A parallel evaluation framework launches multiple expert agents simultaneously, each applying a distinct philosophy and evaluation style to a project. It synthesizes where the experts agree and where they fight, outputting consensus, disagreements, and top questions for decision-making.

Yes, the synthesis block supports targeted director calls, allowing you to request focused feedback from specific expert lenses while still receiving the overall consensus, conflicts, and board verdict from the parallel evaluation.

You need multiple expert lenses when a decision, architecture, or strategy requires multi-perspective review to surface blind spots. It is applicable whenever a user asks for a board's perspective and wants to see where experts agree and where they fight.

No, the directors skill has no dependencies or external components. It operates independently by launching five parallel agents with detailed system prompts and synthesizing their evaluations into a final board verdict.

The best way is to launch separate prompts per AI agent in parallel, then use a synthesis block to report consensus, disagreements, top questions, and recommended actions. This structured approach ensures conflicts are surfaced rather than averaged out.