Mole Debug

What it does

The mole-debug skill allows you to inject debug statements into your AI agent's execution flow. It provides a mechanism to observe internal state and decision-making processes, facilitating troubleshooting and understanding of complex agent behavior. This is achieved by inserting print statements at specified locations within the agent’s code or workflow.

When to use it

• Troubleshooting unexpected behavior: When an AI agent produces incorrect or undesirable results, mole-debug can help pinpoint the source of the problem.
• Understanding decision paths: If you need to understand why an agent made a particular choice, debug statements can reveal the reasoning process.
• Validating internal state: Verify that variables and data structures within your agent are being updated as expected during execution.
• Debugging complex workflows: When dealing with multi-step processes or integrations, mole-debug helps trace the flow of information and identify bottlenecks.

Key capabilities

• Inject debug print statements into code.
• Observe internal state during agent execution.
• Trace decision-making paths.
• Facilitate troubleshooting of complex agent behavior.

Example prompts

• "Add a debug statement before calling the get_weather function."
• "Print the value of user_query after parsing the input."
• "Insert a debug message at line 25 of the main loop to show current iteration."

Tips & gotchas

The skill requires access to the agent's codebase or workflow definition. Be mindful that excessive debugging can slow down execution and potentially impact performance, so remove or disable statements when no longer needed.