Dev Bugger – MERN Stack Starter

This repo contains a full MERN stack scaffold:

• frontend/: Next.js + TypeScript + Tailwind + shadcn-style components
• backend/: Node.js + Express + TypeScript API (dockerized)

Quick start

Frontend

cd frontend
npm install
npm run dev

Backend

cd backend
npm install
npm run dev

Docker (backend)

docker build -t dev-bugger-api ./backend
# or
cd backend && docker compose up --build

Environment variables

Copy the example files and fill values:

• frontend/env.example → frontend/.env.local
• backend/env.example → backend/.env

API Endpoints (Backend)

Base URL: http://localhost:8000/api/v1

Health check

curl http://localhost:8000/api/v1/health

Ingest a file

Ingest a single JSON file (or a directory of JSON files).

curl -X POST \\
  http://localhost:8000/api/v1/ingest \\
  -H "Content-Type: application/json" \\
  -d '{"path":"backend/src/data/daniel_lin_sessions (1).json"}'

Response:

{"status":"ok","path":"backend/src/data/daniel_lin_sessions (1).json"}

Search

Search across embedded segments using vector search.

curl http://localhost:8000/api/v1/search \
  -H "Content-Type: application/json" \
  -d '{"query":"websocket","limit":5}'

Request body fields:

• query (required, q also accepted)
• limit
• username
• projectName
• producer
• producerVersion
• schemaVersion
• taskDescription

Example with filters:

curl http://localhost:8000/api/v1/search \
  -H "Content-Type: application/json" \
  -d '{"query":"logging","username":"daniellin","projectName":"video-streamer","limit":5}'

### AI answer (RAG)

The AI endpoint runs a tool-call to vector search and answers based on results.

```bash
curl http://localhost:8000/api/v1/ai \
  -H "Content-Type: application/json" \
  -d '{"q":"what did andrewwang ask about S3 uploads?","username":"andrewwang","limit":3}'

Sample questions:

• "what did andrewwang ask about S3 uploads?"
• "summarize daniel's conversations about websockets"
• "show Diana's sessions about authentication failures"
• "what changes were suggested for batch encoding performance?"

## Scripts


Backend:
- `dev`: ts-node-dev with watch
- `build`: tsc build
- `start`: run compiled JS
- `pipeline`: run ingestion pipeline via `src/index.ts` (use `--file <path>`)

Example:
```bash
cd backend
npm run pipeline -- --file src/data/daniel_lin_sessions\\ \\(1\\).json

Notes

This scaffold is intentionally minimal. Add your own routes, controllers, and UI pages as needed.