RepoScanAI — GitHub Developer Profile Analyzer

RepoScanAI evaluates GitHub repositories and developer profiles the way a technical recruiter or senior engineer does.

Instead of guessing, it computes structured engineering signals first and then uses AI to interpret them.

It answers practical questions:

Would a company trust this developer in production?

Deployed Project: https://repo-scan-ai.onrender.com/

Video of Demo: https://drive.google.com/file/d/1wEd2QvaBF411AlFPCViR7e80xZlCrjUN/view?usp=sharing

---

What Makes This Different

Most tools analyze code quality.
RepoScanAI analyzes developer credibility.

It performs a multi-layer audit:

• Static Analysis → Repository structure & hygiene
• Behavioral Analysis → Commit patterns & consistency
• Architectural Analysis → System organization
• AI Evaluation → Human-like recruiter judgement

---

Core Evaluation Metrics

The final Professional Grade (A+ → D) is computed from four signals:

1) Structure

Detects engineering discipline.

Checks for:

• README quality
• license
• project organization
• modular separation
• config management

2) Consistency

Measures seriousness of development.

Analyzes:

• commit frequency
• long inactivity gaps
• burst commits (copied projects)
• maintenance behavior

3) Project Depth

Measures real engineering work vs tutorial code.

Looks for:

• multiple modules
• business logic presence
• infrastructure code
• data flow complexity

4) Professionalism

Measures industry readiness.

Detects:

• documentation
• metadata & topics
• naming clarity
• deployability

---

Three Analysis Modes

Mode 1 — Single Repository Audit

Goal: Is this project production-ready?

Provides:

• architecture summary
• strengths & weaknesses
• security risks
• hire / reject verdict

Use cases:

• portfolio review
• project evaluation
• hackathon judging

---

Mode 2 — Repository vs Repository (Peer Comparison)

Goal: Which project shows stronger engineering skill?

Compares:

• architecture (modular vs monolithic)
• maintenance (active vs abandoned)
• complexity (original vs tutorial)
• quality (documented vs messy)

Outputs a winner with reasoning similar to interview panel feedback.

Use cases:

• ranking students
• competitions
• shortlisting candidates

---

Mode 3 — Profile Mode (GitHub Developer Analysis)

Goal: Predict engineering maturity (Beginner → Senior) using measurable signals.

Computes:

• years active
• recent activity consistency
• dominant stack specialization
• serious project count
• impact score (stars + forks)
• domain signals (backend, frontend, ai/ml, devops, data)

Outputs:

• candidate level
• primary stack
• consistency rating
• engineering maturity score
• strengths, weaknesses, recommended roles

---

Features

• AI-generated technical report
• Professional grade scorecard
• Repository comparison
• Developer profile evaluation
• Structured metrics + AI interpretation
• Architecture visualizer
• Security surface detection
• PDF export

---

Tech Stack

Frontend

• HTML5
• TailwindCSS
• Vanilla JS
• Marked.js

Backend

• Node.js
• Express.js
• REST API architecture
• node-fetch

AI Layer

• Google Gemini (gemini-2.5-flash)
• Structured prompt evaluation

---

Local Setup

Requirements

• Node.js ≥ 14
• Google Gemini API Key

Install

git clone https://github.com/Parth-2004/RepoScanAI
cd RepoScanAI/server
npm install

Create .env

GEMINI_API_KEY=your_key_here
PORT=3001

Run

npm start

Open

http://localhost:3001

---

Deployment (Render)

Settings:

Root Directory: server
Build Command: npm install
Start Command: npm start
Environment Variable: GEMINI_API_KEY

After deployment the app automatically serves frontend + backend.

---

Intended Use Cases

• placement preparation
• student portfolio improvement
• hackathon judging
• recruiter screening automation
• open-source contribution assessment

---