common-mistakes-trackerlisted

# Common Mistakes Tracker ## Purpose Maintain a living log of mistakes per project and technology stack. Every time Claude falls into a technical trap, the mistake gets recorded here so it never repeats. This follows Anthropic's internal practice of maintaining a "common mistakes zone" that the model reads before starting work. ## How It Works ### Recording a Mistake When a mistake is identified (by user correction, failed test, or explicit request): 1. Identify the project name or technology 2. Read the existing mistakes log: `mistakes-log.json` in this directory 3. Add the new entry with: - `category`: The tech/framework/language involved - `mistake`: What went wrong (concise) - `correct_approach`: What should have been done - `context`: When this typically happens - `date`: When it was recorded 4. Write the updated log back ### Using the Mistakes Log At the START of any coding task: 1. Read `mistakes-log.json` 2. Filter for entries matching the current technology/framework 3. Keep these in mind as active constraints while coding 4. If about to do something that matches a recorded mistake - STOP and use the correct approach instead ## Mistakes Log Format The file `mistakes-log.json` in this directory stores all entries as: ```json [ { "id": 1, "category": "react", "mistake": "Used useEffect for derived state instead of useMemo", "correct_approach": "Use useMemo for values computed from props/state, useEffect only for side effects",