An AI-powered on-chain behavior analysis tool built on Trac Network / Intercom stack.

Live App: index.html

TRAC AI Wallet Insight Assistant is more than a balance checker — it's a behavioral intelligence layer for TRAC/TAP wallets on Bitcoin.

Users can ask:

• "Analyze my wallet activity"
• "Am I holding or trading too much?"
• "What is my risk profile?"
• "How am I spending my TRAC?"

The AI will:

• ✅ Classify wallet type (Active Trader / Long-Term Holder / Whale / Dormant)
• ✅ Analyze transaction frequency & direction ratios
• ✅ Detect behavioral patterns over time
• ✅ Generate a natural-language insight report
• ✅ Answer follow-up questions via conversational chat

Example AI response:

"You made 18 outgoing transactions this month. Your 72% outgoing ratio classifies you as an Active Trader rather than a long-term holder. Consider batching transfers to reduce network costs."

• Frontend: Pure HTML/CSS/JS — single file, zero dependencies
• AI: Anthropic Claude API (claude-sonnet-4) via Intercom agent integration
• Network: Trac Network / Intercom P2P sidechannels
• Protocol: TAP (Token Allocation Protocol) on Bitcoin

The app features:

• Animated grid background with floating gradient orbs
• Stats row showing Total TXs, Incoming, Outgoing, and Profile type
• AI Insight Report with behavior breakdown bars
• Conversational chat interface

trac14c7k6qdee4hrtr326agwdgy42hkzgwsylymyd3fx3q5zn2phw5ls8xka7n

(Replace this with your actual TRAC address to receive payouts)

1. Clone this repo
2. Open index.html in your browser
3. Enter any TRAC wallet address
4. Click ANALYZE →
5. Ask follow-up questions in the chat

No build step. No server required.

• Upstream Intercom: https://github.com/Trac-Systems/intercom
• Trac Network: https://trac.network
• Awesome Intercom: https://github.com/Trac-Systems/awesome-intercom

MIT