🧬 Docless.ai — The Forensic Code Intelligence Engine
Docless.ai analyzes any public GitHub repository and explains it like a senior engineer — even when there’s zero documentation. It audits architecture, security, activity, and risk so you can decide in minutes, not days, whether a project is worth using.
🚀 Why I Built It
Most repos suffer from the README Gap: outdated docs, no diagrams, and confusing structure. I wanted a tool that treats code like a crime scene — scanning it for hidden signals instead of relying on human explanations.
🛠️ How It Works
Built with aistudio.google.com and Gemini 3 Pro, Docless.ai:
- Maps repository structure recursively
- Extracts key files and entry points
- Generates structured forensic reports (Security, Legal, Performance, Architecture)
Produces an Adoption Readiness Score
(H) Health • (B) Bus Factor • (S) Security • (A) Activity
🚧 Hard Problems Solved
- Huge repos vs. AI limits: structural pruning to fit context windows
- GitHub rate limits: resilient retry layer
- Hallucinations: strict JSON schemas tied to real repo data
🎯 Who It’s For
Developers choosing dependencies, investors auditing tech stacks, and students trying to understand complex projects , with zero code reading required.
Log in or sign up for Devpost to join the conversation.