Hey Product Hunt! 👋 I built Prometheus AI because I was tired of writing the same prompt 3 times before getting a good result. The core insight: a prompt for Claude should look different than a prompt for ChatGPT. They have different strengths, different structures that work best. Same for code vs. marketing copy vs. image generation. So I built a prompt compiler that: • Applies model-specific optimization rules (XML structure for Claude, direct framing for ChatGPT, multimodal clarity for Gemini...) • Layers domain-specific rules based on your use case (10 categories, 60+ subcases) • Keeps your API key in your browser — zero server-side storage It's completely open source (MIT), self-hosted, and deploys anywhere in 2 minutes with no env vars. I'd love your feedback — what use cases would you add? What models should I support next? ⭐ GitHub: github.com/LeDucBSE/prometheus-ai

Prometheus AI

Transform rough prompts into expert-grade AI instructions

Transform rough prompts into expert-grade AI instructions