What problem does it solve?
When a Power Automate flow run fails, manually tracing through run history, action outputs, and flow definitions to find the root cause is slow and error-prone. This Skill automates deep failure analysis of a specific failed run given its environment, flow, and run IDs.
Core Features & Use Cases
- Automated triage: Uses the FlowAgent MCP
diagnose_run tool to classify failed or timed-out actions, then gathers run details, action traces, and the flow definition in parallel.
- Root cause analysis: Builds an execution graph from
runAfter dependencies, distinguishes root failures from cascading skips, and classifies errors such as authorization failures, expression errors, HTTP 4xx/5xx, timeouts, and parameter validation issues.
- Actionable remediation: Produces a diagnosis report with an execution timeline, root cause, specific fix, and confidence level, and can optionally apply the fix directly via
edit_flow or update_flow.
- Use Case: A scheduled flow failed overnight. Provide the environment, flow, and run IDs, and receive a full report identifying the failing action inside a loop, the underlying connection error, and a corrected flow definition.
Quick Start
Ask the agent to diagnose the failed flow run by providing the environment ID, flow ID, and run ID, for example: diagnose flow run with env-id, flow-id, and run-id.
