sadd:do-competitively

What problem does it solve?

This skill enables high-quality task execution by orchestrating multiple agents to generate diverse solutions, evaluate them with multiple judges, and synthesize the best outcome based on evidence.

Core Features & Use Cases

• Self-critique loops in generation (Constitutional AI) to improve ideas.
• Multi-judge evaluation with verification loops to ensure correctness.
• Adaptive strategy selection (polish, redesign, or full synthesis) to balance quality and cost.
• Evidence-based synthesis to combine the strongest elements from parallel approaches.

Quick Start

Provide a task description and constraints, and the orchestrator will run three independent solutions in parallel, evaluate them, and produce a final polished result.