Developers waste time troubleshooting bad prompts for AI coding tools. A validator could score prompts on clarity, specificity and likely effectiveness.
Build a browser extension that analyzes prompts in real-time with color-coded feedback. Add examples of high-scoring prompts.
Sell to engineering teams through team licenses or a freemium model.
Start with basic heuristics for common coding scenarios.
The biggest risk is AI models changing their prompt requirements frequently.