I built a new tool to evaluate and optimize GitHub README files automatically using AI

Hey everyone! 👋
As developers, we all know how crucial a great README.md file is. It’s the front door of any open-source project. But let's be honest—writing, formatting, and scoring a README to make it professional and engaging takes a lot of time and effort.
That’s why I decided to build ReadmeRank, an AI-powered SaaS tool designed to evaluate, score, and optimize GitHub READMEs with smart code validation and instant feedback.
What does ReadmeRank do?
Instant Evaluation: It analyzes your README structure, clarity, and completeness.
Smart Scoring: Gives you a clear score so you know where you stand.
AI Optimization: Provides actionable feedback and suggestions to level up your documentation instantly.
Tech Stack & Architecture
The app is built with Python and Streamlit for a fast, interactive web interface, leveraging AI to handle the heavy lifting of text evaluation.
Check it out & Contribute!
I would love for you to try it out, break it, and share your feedback with me:
Live Demo (Streamlit): Https://readmerank.streamlit.app/
Source Code (GitHub): Https://github.com/am9024048-hue/ReadmeRank
(Note: Since the app is hosted on Streamlit Cloud, it might take a few seconds to wake up if it's inactive—just hit the button to wake it up!)
Let's Connect!
What features would you like to see next? Drop a comment below, leave a star on GitHub if you find it helpful, and let’s discuss how we can make developer documentation better together! 💻✨

Comments

[gmin_step]

How does it actually score "quality" — is it comparing against a fixed rubric (headings, badges, install steps present/missing), or is it an LLM subjectively judging tone and clarity? Asking because those two approaches would fail very differently — a rubric could be gamed by just adding sections without real content, while an LLM judge could penalize valid but unconventional READMEs (e.g. a research repo that's just equations and citations). Curious which tradeoff you picked and why.

[Ahmed]

"Hey! That’s an exceptionally sharp and well-thought-out question.
You hit the nail on the head regarding the tradeoffs of both approaches. To solve this, ReadmeRank leans toward a hybrid strategy:
Structural Check: It ensures essential elements (like installation steps, links, and clear headers) are actually present so it's not completely empty.
Contextual LLM Evaluation: Instead of just looking for empty section headers to game the system, the LLM evaluates whether the content inside those sections actually makes sense relative to the project context.
That said, handling edge cases like pure research repositories (heavy on equations/citations) is a great point, and we are actively fine-tuning the prompt behavior to better adapt to non-standard codebases.
Appreciate you diving deep into the architecture!"