agent-introspection-debugging

To debug an agent stuck in a tool loop, this skill captures the error state and tool usage to diagnose the root cause and apply contained recovery steps to restore progress without broad state mutations.

Agent drift and token drain happen when repeated failures go undiagnosed. This skill captures structured failure states, matches them to known patterns, and suggests minimal remedies to stop the waste.

The best way to capture error states for agent introspection is to record the error type, message, stack traces, last successful step, and relevant context to ground your root-cause analysis.

Yes, you can use structured self-debugging for agent runs in production. It identifies recurring loops and drift from the objective across both development and production contexts to generate an actionable report.

To generate an introspection report after an agent failure, activate the workflow to capture the failure, diagnose the root cause, apply contained recovery, and output a structured human-readable summary of findings and actions.