dispatching-parallel-agents

Dispatch parallel agents by assigning one agent per independent failure domain with a shared wave_id. Each agent investigates its domain independently, then MCP aggregates results into a unified report, reducing total investigation time compared to sequential analysis.

Shannon-wave coordination synchronizes parallel work across multiple domains with quantified progress tracking. It enables structured, measurable parallel execution so each domain agent operates independently while maintaining visibility into overall investigation progress.

Yes. This Skill is designed for multi-domain scenarios where 3+ independent failures occur in distinct subsystems that don't require cross-domain context. Each domain receives its own agent, making it ideal when subsystems fail independently and can be analyzed without shared dependencies.

MCP aggregates per-domain results into a centralized report. Each agent scores its domain investigation (0.00–1.00 success), then MCP collects and consolidates these domain-specific results and scores into a single unified output for review.

Use parallel dispatch when you have 3+ independent failures and need to reduce total investigation time. Sequential investigation is slower; parallel dispatch with per-domain agents and MCP aggregation delivers significant time savings on multi-domain issues without cross-domain dependencies.