I built an AI tool that scores and fixes prompts

Hi Everyone 👋

I’m the solo builder behind PromptFix, a tool I built after noticing that most people struggle to get consistent results from AI because their prompts are vague or poorly structured.

PromptFix analyzes a prompt and returns:

• an overall quality score

• category subscores (clarity, structure, context, constraints)

• specific improvement suggestions

• a rewritten “best version” of the prompt

The goal isn’t to replace prompting — but to help people learn why prompts work or fail.

It connects to my existing backend API and works on web, Chrome extension, and now a mobile version I’m building with Expo.

Some interesting things I learned while building this:

• Most prompts fail due to missing constraints rather than wording

• Users heavily reuse improved prompts once they see structured examples

• Developers tend to want transparency instead of “magic rewriting”

I’d really appreciate feedback from this community — especially around:

• features you’d want as power AI users

• workflows you think are missing

• whether scoring prompts is actually useful or distracting

Happy to answer any technical or product questions.

Thanks 🙏