rubber-duck-brain

graph LR
    A[Have problem] --> B[Start explaining to LLM]
    B --> C[Realize answer while typing]
    C --> D{Send anyway?}
    D -->|Yes| E[LLM suggests same thing]
    D -->|No| F[Solve it yourself]
    E --> G[Feel validated]
    F --> G
    G --> H[Have new problem]
    H --> B

Problem-solving patterns and debugging approaches. The LLM is your rubber duck, except sometimes it quacks back with useful suggestions.

What goes here

• Debugging strategies that work with LLMs
• Problem decomposition patterns
• How to ask questions that get useful answers
• Techniques for getting unstuck
• Templates for explaining complex problems

The rubber duck effect

Half the value of using an LLM for debugging is forcing yourself to articulate the problem clearly. The other half is when it actually solves it. Both are useful.