Agent Introspection Debugging

When your agent burns through tokens in a loop or drifts off task, this gives it a structured way to debug itself before you step in. It walks through four phases: capture the failure state, diagnose the pattern (like repeated tool calls or context bloat), apply a minimal recovery action, and write a legible report. The key discipline is "verify the world state instead of trusting memory" and shrinking scope before retrying blindly. Think of it as teaching the agent to stop, look at what's actually happening, and either fix it with evidence or escalate cleanly. Most useful when you're watching an agent spin its wheels on the same command or slowly lose track of what it was supposed to be doing.

skills.sh installs: 3.7K
GitHub stars: 229.1K
last updated: Jul 14, 2026

View on GitHub

Install

$ npx -y skills add affaan-m/everything-claude-code --skill agent-introspection-debugging --agent claude-code

Installs into .claude/skills of the current project.

Files

Agent Introspection Debugging

Use this skill when an agent run is failing repeatedly, consuming tokens without progress, looping on the same tools, or drifting away from the intended task.

This is a workflow skill, not a hidden runtime. It teaches the agent to debug itself systematically before escalating to a human.

When to Activate

Maximum tool call / loop-limit failures
Repeated retries with no forward progress
Context growth or prompt drift that starts degrading output quality
File-system or environment state mismatch between expectation and reality
Tool failures that are likely recoverable with diagnosis and a smaller corrective action