🌐 Workflows & Swarm
Execution History & Run Traces
View Swarm and agent runs in /executions — filter by status, inspect per-node traces, and export run data.
Updated June 29, 2026
4 min read
Every Swarm canvas run and agent execution is persisted to Execution History. Use it to debug failures, audit costs, and verify inbound triggers (Slack, Discord, webhooks).
What You See
- Status: success, failed, running, pending
- Per-node trace logs with truncated inputs/outputs
- Cost, tokens, latency, provider, and model
- Workflow vs agent source filter
Debugging a Failed Run
1
Open Execution History and filter by Failed.
2
Expand the run to view the trace panel — find the first red node.
3
Check integration credentials in Settings if the node is an integration action.
4
Re-run from Swarm Simulation before deploying again.
Was this helpful?