Hey Product Hunt! 👋 I'm the maker of PromptNote. I write a lot of prompts for ChatGPT, Claude, and Gemini, and I keep running into two problems: 1. My best prompts were scattered across notes, docs, and chat history. 2. I couldn't tell whether a prompt was actually good until I'd wasted a few runs on it. So I built PromptNote, a desktop app that treats prompts like something worth crafting: 🎯 Evaluate: score any prompt on intent clarity, quality, ambiguity, context completeness, and answerability, with concrete suggestions to fix it. ✨ Improve: let AI rewrite the prompt, then compare the result with your original. 🕘 Versions: save every iteration, compare versions, mark one as "Production", and roll back at any time. ⚡ Quick picker: press Ctrl+Shift+P in any app to search and copy a prompt instantly. 🧩 Variables: {{placeholders}} for the parts that change 📁 Collections: Organize prompts, then import and export them to share with your team. 🔌 Bring your own model: OpenAI, Anthropic, Google Gemini, or a fully local Ollama model. Your prompt library lives on your machine. Prompts only go to the provider you connect with, and only when you click Evaluate, Improve, or Run. 💰 Pricing: Try it free for 7 days, then it's a one-time $12 for lifetime access. No subscription. (You bring your own API key, or run a local model for free.) Available now for Windows 10/11. I'd love your feedback, especially: what would make you trust a prompt score? What should the evaluator measure next? Thanks for checking it out! 🙏

PromptNote - Score, improve and version your AI prompts.

Replies