Prompt Engineering

The practice of designing and structuring input prompts to guide large language models (LLMs) toward producing more accurate, relevant, and useful outputs. In accessibility contexts, prompt engineering techniques such as role-play prompting (assigning expert personas), contextual prompting (providing domain-specific context like WCAG guidelines), and metacognitive prompting (structured self-reflection) can significantly improve LLM performance on tasks like detecting and correcting accessibility violations. Research shows that well-engineered prompts can improve LLM correction accuracy from as low as 19% to over 84%.