build-dashboard

name	build-dashboard
description	Builds a dashboard to visualize trades, portfolio, and agent reasoning

When to Use

• When asked to create a visualization interface for the trading agent
• When the user wants to monitor paper trading results
• When building reporting or monitoring features

Instructions

1. Use Streamlit as the primary framework (fallback to Flask if requested)
2. Display the following sections:
   • Portfolio Overview: Current value, total PnL, PnL percentage
   • Trade History: Table of all executed trades with timestamps, prices, sizes
   • AI Prediction Confidence: Chart showing confidence scores over time
   • Strategy Signals: Visual breakdown of signals from each module
   • Risk Manager Warnings: Alerts for rejected trades, limit breaches
   • Proof Log: List of recent validation hashes
3. Update in real-time (auto-refresh every 5 seconds for Streamlit)
4. Use clear, readable charts (line charts for PnL, bar charts for signals)
5. Include filters for date range, strategy type, and trade outcome

Dashboard Layout

┌──────────────────────────────────────────────┐
│  Portfolio Overview                          │
│  Value: $XX,XXX  |  PnL: +$X,XXX (+X.X%)   │
├──────────────────┬───────────────────────────┤
│  PnL Chart       │  Strategy Signals         │
│  (line over time)│  Momentum: BUY (0.72)     │
│                  │  MeanRev:  HOLD (0.45)    │
│                  │  AI Pred:  BUY (0.68)     │
├──────────────────┼───────────────────────────┤
│  Trade History   │  Risk Alerts              │
│  (scrollable     │  - Daily loss: 6.2%/10%   │
│   table)         │  - Max position: OK       │
│                  │  - Volatility: NORMAL     │
├──────────────────┴───────────────────────────┤
│  Validation Proof Log                        │
│  Hash: abc123... | Decision: BUY ETH @ $3200 │
└──────────────────────────────────────────────┘

Data Sources

• Read trade history from simulation/paper_trader.py output
• Read proof hashes from validation/proof_log.jsonl
• Read portfolio state from the paper trader's portfolio tracker
• Read risk alerts from risk/risk_manager.py logs

Example

Input: "Create dashboard for paper trading agent"

Output: dashboard/dashboard.py with:

• Streamlit app with all sections above
• Auto-refresh capability
• Data loading from trade logs and proof logs
• Charts using Streamlit's built-in plotting or matplotlib
• Responsive layout with columns and expanders