⚙️ SkillForge API

Backend service powering SkillForge — an AI-powered learning platform.

This is a REST API built with Express and TypeScript that handles course data, authentication verification, and AI-powered features (chat, recommendations, content generation, and auto-classification) via Groq's LLM API.

🔗 Live API: skill-forge-server-inky.vercel.app 🔗 Frontend Repo: skillforge-web

---

✨ What This API Does

• 📚 Courses — Create, list, search, filter, sort, paginate, and delete course listings
• 📊 Stats & Categories — Aggregated platform statistics and distinct course categories
• 📝 Blog — Serve blog posts and handle contact form submissions
• 🔐 Auth Verification — Verifies JWTs issued by the frontend (better-auth) using JWKS, protecting routes and identifying the authenticated user
• 🤖 AI Endpoints (via Groq):
  • POST /ai/chat & POST /ai/chat/stream — Conversational assistant with streaming (SSE) responses and conversation memory
  • POST /ai/generate — AI content generation with adjustable tone/length
  • GET /ai/recommendations — Personalized course recommendations combining real DB data with AI-generated insights
  • POST /ai/classify — Auto-classifies a course into category, level, and tags

---

🛠️ Tech Stack

• Runtime: Node.js
• Framework: Express.js
• Language: TypeScript
• Database: MongoDB (native driver + Mongoose)
• Auth: JWT verification via jose (JWKS) against better-auth-issued tokens
• AI: Groq, accessed through the OpenAI-compatible SDK (openai package with a custom baseURL)
• File Uploads: Multer
• Validation: express-validator

---

📁 Key Dependencies

Package	Purpose
express	Web server & routing
mongoose	MongoDB object modeling
jose / jwks-rsa / jsonwebtoken	JWT verification against the frontend's JWKS endpoint
openai	Groq API client (OpenAI-compatible)
bcryptjs	Password hashing utilities
multer	Multipart form-data / file upload handling
cors	Cross-origin request handling
express-validator	Request body validation

---

🚀 Getting Started

Prerequisites

• Node.js 18+
• MongoDB (local or Atlas)
• A Groq API key
• The SkillForge frontend running (for JWKS-based auth verification)

Installation

git clone https://github.com/<your-username>/skillforge-api.git
cd skillforge-api
npm install

Environment Variables

Create a .env file in the root:

PORT=5000
MONGODB_URI=your_mongodb_connection_string
DB_NAME=skillforge
NEXT_PUBLIC_BASE_URL=http://localhost:3000
AI_API_KEY=your_groq_api_key
AI_MODEL=openai/gpt-oss-20b

NEXT_PUBLIC_BASE_URL must match your running frontend's URL — it's used to fetch the JWKS public keys for verifying auth tokens.

Run locally

npm run dev

The API will be available at http://localhost:5000.

Build for production

npm run build
npm start

---

📡 API Overview

Method	Endpoint	Description
GET	/api/v1/courses	List courses (search, filter, sort, paginate)
GET	/api/v1/courses/:id	Get a single course
POST	/api/v1/courses	Create a course (auth required)
DELETE	/api/v1/courses/:id	Delete a course (auth required)
GET	/api/v1/courses/categories	Get distinct categories
GET	/api/v1/courses/stats	Platform statistics
GET	/api/v1/blogs	List blog posts
GET	/api/v1/blogs/:slug	Get a single blog post
POST	/api/v1/blogs/contact	Submit a contact form
POST	/ai/chat	AI chat (non-streaming)
POST	/ai/chat/stream	AI chat (SSE streaming)
POST	/ai/generate	AI content generation
GET	/ai/recommendations	AI-powered course recommendations
POST	/ai/classify	AI course auto-classification

---

📄 License

This project is open source and available under the MIT License.

---

Built as the backend counterpart to SkillForge — demonstrating secure API design, JWT-based auth verification, and practical agentic AI integration.