Orchestrate a goal-owning agent + specialist subagents + an eval gate to produce and verify complex output. Use when a single-agent loop can't hold the full scope — strategy, multi-surface campaigns, large audits.
Orchestrate a goal-owning agent + specialist subagents + an eval gate to produce and verify complex output. Use when a single-agent loop can't hold the full scope — strategy, multi-surface campaigns, large audits.