prompt-engineer

Optimize and test LLM prompts for the research agent. Use when writing prompts, improving prompt quality, debugging LLM responses, fixing JSON parsing issues, or tuning output format.