MSAD — Mishra & Sheth Analysis Dashboard

A beginner-friendly stock analysis dashboard by Aastik Mishra and Rahil Sheth, built with Next.js. Browse live market pulse, open guided report cards for any ticker, and learn fundamentals, charts, news sentiment, options, and modeling — without buy/sell advice.

Educational only — not financial advice.

Features

Learn, test, decide — the part no other market app has

MSAD is a trading gym, not a tip service. The goal is to make you a measurably better investor, then let you carry that earned conviction to the real market.

• The Lens — a grounded AI tutor on every stock. One depth slider (Learn → Analyst → Quant) re-explains the same real numbers at your level. Ask anything, or hit "Make the bear case" to fight your own bias. Never gives buy/sell advice; answers are anchored to the stock's actual data so it can't make figures up.
• The Conviction Journal — write a thesis (and what would change your mind) before you act. The AI coach critiques your reasoning, not the outcome.
• Calibration score — log probability predictions; we resurface each on its due date so you score yourself. Your readiness score is the honest "am I getting better?" signal that bridges practice to real money.
• Hypothesis Lab (/practice) — backtest a rule (buy & hold, or a moving- average trend filter) against real EOD history vs buy-and-hold and the S&P 500. Every result ships with the honest caveats — survivorship bias, overfitting, costs, no look-ahead — so a backtest can't fool you.
• Market Replay (/replay) — trade a real, anonymized stretch of history one day at a time, blind to the future, then see your timing vs buy-and-hold. The best cure for hindsight bias.
• Portfolio Risk X-ray (/risk) — paste your holdings (or import them read-only from a linked brokerage via SnapTrade) and see hidden concentration, sector tilt, correlation, and a beta-based estimate of what a market crash could cost you. We never place trades.
• Your dashboard (/dashboard) — journal, predictions, calibration, and watchlist in one place. Saved locally as a guest; synced to your Google account across devices when signed in.

Home & discovery

• Netflix-style catalog with live quotes, sparklines, and sector heatmap
• Market news feed and Market Pulse ticker strip
• Discover screener with beginner presets, Pro filters, saved screens, and exclusions
• Explore — filter the full universe by sector, exchange, cap, and fundamentals
• Contour landscape backdrop tinted by today's market mood

Report cards

• Live data from Financial Modeling Prep (profile, quote, ratios, peers, history)
• Overall A–F grade with strengths / watch-outs
• Seven metrics vs industry peers with Learn / Pro explainers
• Guided learning path on each stock page
• News headlines, analyst sentiment, and fundamentals tension
• Interactive charts: 1W–MAX, compare, scatter, volume, PNG export
• /compare — side-by-side grades, metrics, verdict, chart overlay

Options & modeling

• Options lab: chain, greeks, strategies, what-if sliders, custom builder
• Calculators: position size, breakeven, covered call, DCA
• Monte Carlo fan chart and peer-implied fair value

Personalization

• Watchlist with live refresh and sort by % change
• Dark / light mode, accent color, optional UI sounds
• First-visit onboarding
• Keyboard shortcuts: / search, 1–7 metrics

Export & share

• Report PNG, story snapshot, share links with chart state

Local development

npm install
cp .env.example .env.local   # add FMP_API_KEY
npm run dev

Open http://localhost:3000.

API routes

Route	Description
GET /api/report/{TICKER}	Full report card JSON
GET /api/chart/{TICKER}	Chart payload (compare / scatter)
GET /api/options/{TICKER}	Options chain payload
GET /api/quotes	Catalog quotes; ?symbols=AAPL,MSFT for watchlist
GET /api/sparklines?symbols=	Mini sparkline arrays
GET /api/news/market	Market news feed
GET /api/news/{TICKER}	Ticker news
GET /api/earnings/{TICKER}	Next earnings date
POST /api/screener	Stock discovery / screener
POST /api/explore	Explore universe filters
POST /api/explain	The Lens — grounded AI explanation at a depth
POST /api/backtest	Hypothesis Lab — backtest a rule vs buy-hold & SPY
GET /api/replay	Market Replay — a random anonymized price window
POST /api/risk	Portfolio Risk X-ray — concentration, beta, crash math
GET /api/brokerage/status	Whether brokerage import is available/linked
POST /api/brokerage/link	Start a read-only SnapTrade brokerage link
GET /api/brokerage/holdings	Imported real holdings (weights by ticker)
GET/PUT /api/profile	Signed-in user's cloud progress
GET /api/auth/session	Current user + whether auth is enabled
GET /api/auth/google	Start Google sign-in (when configured)

Deploy on Vercel

1. Push to GitHub
2. Import in Vercel
3. Set environment variables (see .env.example):
   • FMP_API_KEY — market data (required for live data)
   • ANTHROPIC_API_KEY — the AI Lens (optional; static fallback without it)
   • MASSIVE_API_KEY — news + sentiment (optional)
   • GOOGLE_CLIENT_ID / GOOGLE_CLIENT_SECRET / AUTH_SECRET — accounts (optional)
   • SNAPTRADE_CLIENT_ID / SNAPTRADE_CONSUMER_KEY — brokerage import (optional)
   • KV_REST_API_URL / KV_REST_API_TOKEN — durable cross-device sync (optional)
4. Deploy

Every integration degrades gracefully: with no keys at all the app still runs on sample data, curated explainers, and local (guest) progress.

Project structure

src/app/                 Pages & API routes
src/components/          UI (dashboard, charts, options, news)
src/lib/aggregator/      FMP → ReportCard pipeline
src/lib/screener/        Discovery filters, presets, saved screens
src/lib/news/            News aggregation and sentiment
src/lib/options/         Black-Scholes chain & strategies
src/lib/brand.ts         MSAD name, storage keys, events

Scripts

Command	Description
npm run dev	Dev server
npm run build	Production build
npm run start	Start production server