Debug

What it does

The inkeep-debug skill provides debugging capabilities for AI agents, allowing users to inspect and understand the agent's internal state. It facilitates identifying issues in complex workflows by providing access to intermediate variables and process steps. This helps pinpoint errors and improve overall agent performance.

When to use it

• Troubleshooting unexpected behavior: If your agent isn’t performing as expected, use this skill to examine its thought process.
• Understanding complex logic: When dealing with intricate decision-making processes within the agent, debug helps visualize the flow and identify bottlenecks.
• Validating intermediate results: Verify that the agent is producing correct outputs at each stage of a multi-step task.
• Debugging custom tools or functions: When integrating new tools or functions into your agent’s workflow, use this to ensure they are operating correctly.

Key capabilities

• Inspect internal variables
• Trace process steps
• Identify errors in workflows
• Visualize agent thought processes

Example prompts

• "Debug the last step of my agent's reasoning."
• "Show me the values of all variables after this function call."
• "What was the intermediate result before the final decision?"

Tips & gotchas

This skill is most effective when integrated within a structured agent environment. It requires careful prompt engineering to target specific debugging points and interpret the output effectively.