Reflection: Prompt Engineering Practices

Reflect on prompt engineering practices used in this RAG system.

Current Prompts:
{{prompt_templates}}

Performance Data:
{{prompt_performance}}

User Feedback:
{{user_feedback}}

Reflection Questions:
1. Which prompts consistently produce high-quality outputs?
2. Where do prompts fail to guide the model effectively?
3. How well do prompts handle edge cases?
4. Are prompts efficiently using the context window?
5. What patterns from successful prompts can be generalized?
6. How should prompt versioning and testing improve?

Provide actionable improvements for each prompt template.