Adversarial convergence decision

Converge multi-agent swarm decisions

Converging multiple specialized agents into a unified swarm decision

• · { agents[], topic, decision_schema }

• · { consensus_verdict, agent_votes[], rationale }

• · multi-agent workflow execution
• · parallel task coordination
• · workflow phase tracking

• · schema validation framework
• · optional arguments object

• · failed results filtered silently

• · explicit StructuredOutput schemas
• · adversarial verification pattern
• · parallel barrier for race detection
• · explicit verdict field
• · confidence scoring
• · named phase tracking
• · null-filtering for fault tolerance