6.2
Module 6 · Multi-Agent Orchestration
Routing, Handoffs & Failure Modes
Design routing rules between specialist agents, simulate 30 requests through the system, and watch how failures cascade or get absorbed.
Handoff Simulator
Failure Cascade Visualiser
Handoff Simulator
Configure your specialist agents and their capabilities, then run 30 simulated requests to see how routing, handoffs, and retries play out in practice.
Specialist Agents
0
Succeeded
0
Retried
0
Failed
Key insight: The retry policy you choose dramatically affects system reliability. "Retry once" catches most transient failures, but "fallback agent" is the only strategy that handles persistent capability gaps gracefully.
Failure Cascade Visualiser
Click any agent node below to inject a failure. Watch how the failure propagates through the chain -- does the system recover, retry, or cascade the failure to downstream agents?
Agent Chain (click a node to inject failure)
Click an agent node above to simulate a failure at that point
Recovery strategy: