📊 TradeInsightsAssistant

Conversational AI for Options Intelligence

🚀 What is TradeInsightsAssistant?

TradeInsightsAssistant is your options research analyst — powered by natural language. It transforms open interest (OI), volatility positioning, and cross-asset flows into precise trading insights using a powerful chat interface backed by LLMs.

Ask it questions like:

• "Where are traders piling on SPY puts this week?"
• "What’s the risk/reward on a June QQQ put spread?"
• "Are there unusual builds in NVDA for Friday’s expiration?"

🧠 Key Capabilities

💬 Conversational Intelligence

• Ask questions in plain English
• Receive clean, structured responses
• Get Markdown-formatted trade setups

📈 Options Market Analysis

• Open Interest clustering (per strike and expiry)
• Call/Put ratio skew detection
• Max Pain and magnet zones
• Position buildup and liquidation tracking

⚒️ Professional Trade Outputs

• Spreads, hedges, directional trades
• Entry, stop-loss, profit target rules
• Kelly Criterion sizing and confidence score

🛠️ How It Works

1. Launch the CLI: python main.py
2. Choose your LLM backend (Bedrock or Claude API)
3. Start chatting:

📊 You: What are key OI levels on TSLA this week?
🤖 Agent: The 280 and 300 put strikes show growing OI with strong skew. Gamma risk is concentrated near 290.

4. Want to save the response? Just say yes — it writes to Markdown with a full metadata trail.

🔍 Use Cases

• 0DTE and weekly trade planning
• Institutional flow tracking
• Quantified trade generation
• Hedge scenario simulation
• Risk-layered market analysis

⚡ Example Prompts

"Analyze SPY open interest for the next 5 days"
"What’s the options flow telling us about AAPL?"
"Show me support and resistance levels for TSLA"
"Give me a put hedge strategy for HOOD next month"

Run the CLI:

python main.py

You’ll be guided to choose your LLM provider (AWS Bedrock or Claude API). From there, start chatting.

📂 MCP Server Configuration

Servers are defined in mcp_servers.json:

{
  "servers": [
    {
      "name": "openinterest",
      "command": "mcp-openinterest-server",
      "args": []
    },
    {
      "name": "news",
      "command": "mcp-news-server",
      "args": ["--api-key", "your-key"]
    }
  ]
}

Add/remove servers as needed — the CLI dynamically loads them.

🧩 Architecture

• ✅ CLI-first UX built on rich
• ✅ Orchestrator handles multi-turn LLM conversations
• ✅ Dynamic tool invocation from MCP registry
• ✅ Markdown + JSON output for traceability

📜 License

MIT License