metrics-analysis

Root cause analysis from task metrics and error patterns uses 5-Why reasoning and an error taxonomy covering IMPORT, ASSERT, RUNTIME, DB, and TYPE categories to pinpoint actionable insights for performance improvements.

Analyzing error taxonomy data for prompt evolution uses a canonical fitness function and structured metric formats to evaluate performance monitoring data and generate targeted improvement suggestions for the Evolver workflow.

To start analyzing metrics and generating prompt improvement suggestions, run the Analyzer to load task execution data, apply 5-Why root-cause reasoning, and output structured formats required by the Evolver.

Yes, this metrics analysis approach supports runtime and database errors because the error taxonomy explicitly categorizes RUNTIME and DB failures alongside IMPORT, ASSERT, and TYPE errors for comprehensive root-cause identification.

You need structured metrics and a fitness function for root-cause analysis when automating problem identification in task performance monitoring and preparing prompt evolution workflows that require canonical data formats.