A domain-agnostic performance optimization system applies first principles from an external knowledge base to domain structure specifications to derive objectively valuable characteristics. A neural network performs simulation-based analysis identifying which agent characteristics enable constraint satisfaction. The system generates metric profiles quantifying individual agent capability levels across the objectively valuable characteristics at a granular level, capturing capability heterogeneity. The system performs three-way optimization matching tactical sequences to agent capabilities under operating conditions by simultaneously considering tactical sequence optimality, capability requirements for execution, available agent capabilities, and current operating conditions. The system operates domain-agnostically through consistent analytical frameworks across domains, with domain adaptation occurring through domain structure specification, data acquisition adapted to available sources, and output interpretation translating results into domain-specific semantic meanings.