# How to Vibe Code Your Own TradingWizard (and Stop Paying for It) > AI that trades with you across 135+ global markets - Site: https://tradingwizard.ai - Category: Fintech / AI Trading - Verdict: **Serious undertaking** (38/100 vibecodeable) - Estimated effort: 4-6 weeks of focused development ## Verdict You can build a functional single-user paper trading dashboard and LLM market assistant, but orchestrating real-time feeds across 135+ markets and a bidirectional MCP server requires weeks of intense engineering. The core UI of a crypto/stock dashboard paired with a chat assistant is straightforward to scaffold with Claude Code and Next.js. However, the real friction lies in building the backend polling worker infrastructure that ingests live multi-asset prices every few seconds, evaluates trading strategies deterministically, simulates paper-trade execution accurately without a real broker hook, and exposes a fully compliant Model Context Protocol (MCP) server with mixed public/OAuth scopes. Unless you are building this specifically to master real-time streaming architectures and agentic protocol design, subscribing to the real product is vastly more efficient than wrestling with live market data normalization. ### What you can't replicate - The exact historical backtested strategy data and proprietary bot confidence weightings - Curated institutional Deep Research report library ## What it does TradingWizard is an AI-powered market analysis and paper-trading platform designed for independent traders, functioning as an automated research desk and strategy sandbox. ### Core features - Autonomous paper-trading bots scanning live market feeds - Wiz AI conversational assistant with multi-source context - Model Context Protocol (MCP) server endpoint for external AI clients - SEC Form 4 insider trading tracker and economic calendar - Read-only exchange portfolio health sync (Binance/Bybit) - Deep Research multi-source report generation desk ## The business ### Pricing - Starter: Free — For getting a feel for Wiz - Pro: $39/mo — For traders who want the full desk - Ultimate: $99/mo — For traders running bots on every market Founded 2026. Team size: 1. ## The hard parts - Normalizing disparate live feeds across crypto, stocks, forex, and futures at 5-8 second polling intervals - Orchestrating a mixed-auth MCP server supporting public read-only tools and OAuth-protected private actions - Building a deterministic paper-trading execution engine with reliable stop-loss, take-profit, and slippage simulation - Managing long-running, token-heavy Gemini reasoning loops for complex multi-source technical analysis ## How to vibe code TradingWizard ### Prerequisites - Node.js (free): Required runtime for the Next.js framework and backend workers. - GitHub (free): Source control and deployment pipeline integration. ### Recommended AI tools - Claude Code: Best-in-class agentic coding tool for scaffolding full-stack Next.js architectures, MCP servers, and background cron workers. - Cursor: Ideal editor for fine-tuning complex React dashboard components, charting UI, and real-time state hooks. ### Stack - Frontend: Next.js with Tailwind CSS and Shadcn UI components - Backend: Next.js API Routes & Server Actions / Node.js background workers - Database: Neon (Serverless Postgres) - Auth: better-auth - Payments: Stripe - Other: Vercel AI SDK, Google Gemini API (via Gemini API), Resend ### Hosting - Vercel (Hosting the Next.js frontend, API endpoints, and serverless cron tasks): $0-20/mo - Neon (Storing user portfolios, bot configurations, and paper trade logs): $0/mo ### Build guide 1. **Project Scaffolding & Database Schema** — Initialize the Next.js project with Tailwind CSS, configure better-auth with Neon PostgreSQL, and set up database schemas for users, bots, watchlists, paper trades, and market snapshots. ``` Scaffold a new Next.js project using App Router, TypeScript, and Tailwind CSS. Integrate better-auth with a Neon PostgreSQL connection using Drizzle ORM or Prisma. Define database tables for users, bot_configurations (storing symbol, strategy, risk limits), bot_scans (logging timestamps, plain-language reasoning, score), and paper_trades (side, entry price, stop-loss, take-profit, P&L status). Ensure environment variables are cleanly structured and include error boundaries for database connection failures. ``` 2. **Live Market Ingestion & Watchlist Dashboard** — Build the core dashboard UI displaying global markets, watchlists, and live price tickers with polling or WebSocket mechanisms. ``` Build a responsive trading dashboard UI using Tailwind CSS and Lucide icons that mimics a pro terminal layout. Implement a watchlist sidebar supporting stocks, crypto, and forex assets. Create API routes that fetch live price feeds from public financial APIs (or mocked data endpoints for personal use if rate-limited), displaying current prices, 24h change, and pre-market status. Build client components that poll or update prices cleanly every 8 seconds. ``` 3. **Autonomous Paper-Trading Bot Engine** — Develop the automated background worker loop that periodically evaluates active bot strategies against incoming market candles and executes simulated paper trades. ``` Implement a backend background worker or scheduled serverless cron job in Next.js that iterates through active user bot configurations. For each bot, fetch recent candlestick and technical indicator data, invoke the Google Gemini API using the Vercel AI SDK to score the setup, generate a plain-language rationale, and calculate entry, stop-loss, and take-profit targets. If setup criteria are met and no position is open, insert a simulated paper trade record into the database and log the scan receipt to the bot timeline. ``` 4. **Wiz AI Assistant & Chat Terminal** — Create the conversational assistant interface with tool-calling capabilities to query market conditions, portfolio risk, and bot status. ``` Build an AI chat interface component ('Wiz') using the Vercel AI SDK on the frontend and backend. Implement server-side tools that allow the AI model to query user bot statuses, check current portfolio risk, fetch live market prices, and summarize recent SEC Form 4 insider filings. Ensure the UI supports markdown rendering, streaming responses, and displays tool-call execution steps clearly. ``` 5. **Model Context Protocol (MCP) Server Endpoint** — Expose a Model Context Protocol (MCP) server endpoint enabling external LLM clients like Claude Code and ChatGPT to query market tools and manage bot sessions. ``` Implement a Model Context Protocol (MCP) server endpoint at `/api/mcp` using a standard TypeScript MCP adapter. Structure the endpoint to support mixed authentication: public read-only tools (fear & greed, market movers, global bot reads) accessible without authentication, and private account tools (user bots, paper positions, portfolio risk) protected via OAuth token verification. Register at least 10 core tools matching the spec and verify they respond correctly to external MCP client requests. ``` 6. **SEC Form 4 Insider Tracker & Deep Research Desk** — Add macro intelligence feeds including SEC Form 4 insider trading logs and an AI-driven research report generator. ``` Build a dedicated 'Who's Moving Markets' panel that ingests and displays recent SEC Form 4 insider trading filings, mapping executive buy/sell activities to tickers in the user's watchlist. Implement a 'Deep Research' generator page where users can request institutional-style markdown reports on any ticker or crypto asset, utilizing the Gemini API with search grounding to synthesize thesis, key numbers, risks, and verified citations. ``` 7. **Portfolio Sync & Polish** — Add read-only portfolio integration for exchange accounts, comprehensive trade log history, and final UI polish. ``` Implement a portfolio sync settings page allowing users to securely input read-only API keys for simulated or real exchange tracking (e.g., Binance/Bybit). Build a comprehensive trade log table with win-rate calculations, P&L percentages, and filterable history by bot and symbol. Refine mobile responsiveness and ensure clean error handling across all API routes. ``` ### Cost vs paying **Starting costs (one-time):** - Domain name (optional): $12 one-time - Total: ~$12 one-time **Ongoing costs (monthly):** - Vercel Hobby / Pro Hosting: $0-20/mo - Gemini API usage for bot scans & chat: $5-15/mo - Neon Database: $0/mo - Total: ~$15-35/mo - Paying for the SaaS instead: $39/mo (Pro) or $99/mo (Ultimate) - Build time: 40-60 hours - AI tool credits: $20/mo (Claude Pro) - Break-even: 1-2 months of coding time vs subscription cost if valued at hobby rates, though building takes significant effort. ## Sources - [TradingWizard Homepage](https://tradingwizard.ai) - [TradingWizard Pricing](https://tradingwizard.ai/pricing) - [TradingWizard Features](https://tradingwizard.ai/features) - [TradingWizard About](https://tradingwizard.ai/about) - [TradingWizard MCP Documentation](https://tradingwizard.ai/docs)