Skip to main content
Glama
DEMO_SCRIPT.md5.26 kB
# Demo Video Script ## Overview **Duration**: 3-4 minutes **Tone**: Professional but approachable, enthusiastic **Format**: Screen recording with voiceover --- ## Script Outline ### Scene 1: Landing Page (20 seconds) **Visual**: MonteWalk landing page **Voiceover**: > "Meet MonteWalk—an institutional-grade quantitative trading platform built for AI agents. Retail traders struggle with data overload and lack professional risk tools. MonteWalk solves this by connecting AI assistants like Claude to real-time market data, portfolio management, and advanced analytics—all through the Model Context Protocol." **Action**: Scroll slowly through feature cards --- ### Scene 2: Dashboard Tour (40 seconds) **Visual**: Click "Test via UI" button, navigate to Dashboard **Voiceover**: > "The Gradio 6 interface gives you a professional trading terminal in your browser. The dashboard shows your portfolio holdings, watchlist with live prices, trending cryptocurrencies, and aggregated news—all in one view." **Action**: 1. Show portfolio panel (point out cash, positions) 2. Highlight watchlist with prices 3. Scroll through crypto trends 4. Show news headlines 5. Click "Refresh Data" to demonstrate live updates --- ### Scene 3: Toolbox Deep Dive (50 seconds) **Visual**: Navigate to Toolbox tab **Voiceover**: > "Behind the scenes, MonteWalk provides 25 specialized tools across 9 categories. Let's test a few." **Action**: 1. **Market Data**: Open `get_price`, enter "AAPL", period "1mo", show chart 2. **Technical Analysis**: Open `get_technical_summary`, enter "AAPL", show BUY/SELL signal with reasoning 3. **Risk Management**: Open `portfolio_risk`, show volatility calculation 4. **News & Sentiment**: Open `get_symbol_sentiment`, enter "TSLA", show FinBERT analysis **Voiceover continues**: > "Each tool is production-ready with error handling, multi-source fallbacks, and intelligent caching." --- ### Scene 4: MCP Integration (40 seconds) **Visual**: Navigate to "Setup MCP Client" page **Voiceover**: > "The real power comes from connecting MonteWalk to your AI assistant via MCP. Claude Desktop, VSCode, Cursor—any MCP-compatible client can access these tools." **Action**: 1. Show Claude Desktop config snippet 2. Highlight the simple setup (2 lines of JSON) **Visual**: Switch to Claude Desktop window (if available) **Voiceover**: > "Once connected, your AI can orchestrate complex workflows—like this morning briefing that analyzes portfolio risk, scans for unusual activity, and summarizes top news." **Action**: Type in Claude: "Run morning briefing" (show output if possible) --- ### Scene 5: Agentic Workflows (30 seconds) **Visual**: Back to MonteWalk UI, show Prompts accordion **Voiceover**: > "MonteWalk includes pre-built agentic workflows. The Morning Gamma Hunt finds unusual market activity, deep-dives the top 3 picks, and ranks them by conviction score. The Portfolio Rebalance workflow calculates optimal weights using modern portfolio theory." **Action**: 1. Expand Prompts accordion 2. Show `morning_gamma_hunt` prompt text 3. Show `portfolio_rebalance` prompt --- ### Scene 6: Technical Highlights (20 seconds) **Visual**: Screen showing code editor with app.py **Voiceover**: > "Built with Gradio 6's native MCP support, custom theming, and a robust data pipeline that aggregates from Yahoo Finance, Alpaca, CoinGecko, and NewsAPI—with FinBERT sentiment analysis running on Modal's serverless GPUs." **Action**: Quick scroll through app.py structure --- ### Scene 7: Real-World Impact (15 seconds) **Visual**: Back to landing page, scroll to "Real-World Impact" section **Voiceover**: > "MonteWalk democratizes institutional-grade tools for retail traders, provides a reference implementation for financial MCP servers, and showcases the power of agentic AI in finance." --- ### Scene 8: Call to Action (15 seconds) **Visual**: Gradio UI with clear URL visible **Voiceover**: > "Try MonteWalk on Hugging Face Spaces, connect it to Claude Desktop, or run it locally. Links in the description. Built for the MCP 1st Birthday Hackathon—where AI meets quantitative finance." **Action**: Show GitHub star button, HF Space link --- ## Recording Tips ### Technical Setup - **Resolution**: 1080p minimum - **Tool**: OBS Studio, Loom, or QuickTime - **Audio**: Clear microphone, no background noise - **Speed**: Speak at moderate pace, pause between sections ### Visual Quality - **Hide Distractions**: Close unrelated tabs, clear desktop - **Highlight**: Use cursor highlights or zoom for important UI elements - **Smooth Transitions**: No jarring cuts, use fade when switching contexts ### Content Focus - **Show, Don't Tell**: Let the UI speak for itself - **Real Data**: Use actual API responses, not mocked data - **Enthusiasm**: Convey genuine excitement about the project --- ## Post-Production Checklist - [ ] Add text overlays for key features - [ ] Include background music (low volume, non-distracting) - [ ] Add title card with "MonteWalk - MCP 1st Birthday Hackathon" - [ ] Include end card with links: - HF Space URL - GitHub URL - MCP Hackathon page - [ ] Export in MP4 (H.264) - [ ] Upload to YouTube (unlisted is fine) - [ ] Embed in README

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/N-lia/MonteWalk'

If you have feedback or need assistance with the MCP directory API, please join our Discord server