Dashboard Service

Node.js backend that powers a personalized crypto dashboard. It ingests live market data, fetches curated news, generates AI-based insights or memes, and stores end-user preferences so a frontend can render dynamic dashboards. The service exposes both REST and WebSocket interfaces and is built with Express, Socket.IO, Axios, and Mongoose.

Features

• User profile CRUD - persist investor preferences via MongoDB so AI and content responses can be contextualized.
• AI insights - wrap Google Gemini through @google/genai to generate tailored guidance or meme prompts.
• News aggregation - fetch and normalize CryptoPanic articles for downstream consumption.
• Crypto market data - hydrate dashboards with CoinGecko spot prices through both REST (/api/crypto) and a streaming Socket.IO channel.
• Meme discovery - search meme/GIF APIs directly or derive prompts from the latest news cycle.
• CORS-friendly gateway - single server that enables both REST endpoints and WebSocket events for any frontend origin during development.

Architecture

server.js
|- routes/ (REST surface mounted at /api)
|  |- userDataRoutes.js -> wires controllers
|- controllers/ (request validation/orchestration)
|  |- userDataController.js  -> CRUD + query helpers
|  |- aiController.js        -> Gemini-backed responses
|  |- newsController.js      -> CryptoPanic fetcher
|  |- memeController.js      -> Meme discovery
|  |- cryptoController.js    -> CoinGecko proxy
|- services/ (API integrations & utilities)
|  |- aiService.js           -> Gemini helper + prompt builder
|  |- cryptoService.js       -> CoinGecko wrapper
|  |- memeService.js         -> Meme API client
|  |- newsService.js         -> CryptoPanic client
|- sockets/cryptoSocket.js   -> `prices` Socket.IO namespace
|- models/UserData.js        -> Mongoose schema

server.js configures Express middleware (body parsing, CORS), mounts /api routes, registers error handling, and instantiates Socket.IO on the same HTTP server. The socket layer polls CoinGecko every INTERVAL_MS (default 120s) for whichever symbol list the client last provided and emits updates via the prices channel.

Prerequisites

• Node.js 20+ and npm
• Running MongoDB instance reachable through DB_URI
• API credentials:
  • Google Gemini (GEMINI_API_KEY)
  • CryptoPanic (CRYPTOPANIC_API_KEY)
  • Meme / GIF provider (MEME_API_KEY)
  • CoinGecko (COINGECKO_API_KEY)

Environment Variables

Create a .env at the repo root (never commit production secrets). Example:

PORT=5002
DB_URI=mongodb://localhost:27017/dashboard-service
CLIENT_ORIGIN=http://localhost:4200
GEMINI_API_KEY=your-google-gemini-key
AI_MODEL=gemini-2.5-flash
CRYPTOPANIC_API_KEY=your-cryptopanic-token
CRYPTOPANIC_BASE_URL=https://cryptopanic.com/api/v1/posts/
MEME_API_KEY=your-meme-api-key
MEME_API_BASE_URL=https://api.apileague.com/search-gifs
COINGECKO_API_KEY=your-coingecko-key
COINGECKO_API_BASE_URL=https://api.coingecko.com/api/v3/simple/price

Note: server.js currently expects MongoDB to be connected before CRUD routes are hit. Add mongoose.connect(process.env.DB_URI) somewhere during bootstrap or ensure another module initializes the connection.

Installation & Local Development

npm install
# Optionally add nodemon for auto-reload (already listed as dependency)
npx nodemon server.js   # or: node server.js

The server logs Server listening on port <PORT> once ready. REST endpoints default to http://localhost:5002/api/....

API Reference

Method	Path	Description
POST	/api/userdata	Create a profile (name, email, optional assets, investorType, contentType[]).
GET	/api/userdata	List all stored user documents.
GET	/api/userdata/:id	Fetch a profile by Mongo _id.
GET	/api/userdata/email/:email	Fetch a profile by email.
PUT	/api/userdata/:id	Update profile fields.
DELETE	/api/userdata/:id	Remove a profile.
POST	/api/ai	Send { "prompt": "..." } and receive AI-generated text.
POST	/api/aiInsights	Body combines user preferences + optional overrides; endpoint pulls CryptoPanic news and returns AI-crafted insights.
GET	/api/news	Normalized CryptoPanic feed.
POST	/api/meme	Provide { "title": "bear market" } to fetch a matching meme URL.
POST	/api/memeAI	Provide user context; service will build a prompt from live news, get AI keywords, then fetch a meme URL.
GET	/api/crypto	Proxy to CoinGecko, returning usd quotes and 24h change for IDs defined in cryptoService.

Sample: create a user profile

curl -X POST http://localhost:5002/api/userdata \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Casey",
    "email": "casey@example.com",
    "assets": ["btc", "eth"],
    "investorType": "long_term",
    "contentType": ["insights", "memes"]
  }'

Sample: request insights

curl -X POST http://localhost:5002/api/aiInsights \
  -H "Content-Type: application/json" \
  -d '{ "assets": ["btc"], "riskTolerance": "medium" }'

The controller will pull latest CryptoPanic posts, enrich the prompt with your payload, and return Gemini text.

WebSocket Stream

• Endpoint: ws://localhost:5002 (Socket.IO)
• Client event: set_crypto_ids -> body can be ["bitcoin","ethereum"] or "bitcoin,ethereum"
• Server emits:
  • prices - payload straight from CoinGecko ({ bitcoin: { usd: 42000, usd_24h_change: -1.2 }, ... })
  • prices_error - indicates validation or fetch failures

import { io } from "socket.io-client";
const socket = io("http://localhost:5002");
socket.emit("set_crypto_ids", ["bitcoin", "solana"]);
socket.on("prices", (data) => console.log(data));
socket.on("prices_error", (err) => console.error(err));

Testing

There are no automated tests yet (npm test exits intentionally). For now rely on manual calls (cURL, Postman, or the included client.html) and consider adding integration tests around the REST controllers when time permits.

Troubleshooting

• 401/403 from external APIs - verify each provider key is present and has remaining quota.
• Mongo validation errors - make sure name and email are provided and unique.
• Socket emits nothing - you must emit set_crypto_ids at least once; also ensure CoinGecko rate limits are not exceeded.
• CORS issues - override CLIENT_ORIGIN in .env with your frontend URL when moving past local development.

Next Steps

1. Add a startup script (npm run start) and wire up the missing mongoose.connect.
2. Implement request validation (e.g., Zod/Joi) before controllers hit external APIs.
3. Add caching for CryptoPanic responses when their rate limit becomes restrictive.