Build me a UAE fuel price forecast dashboard that predicts next month's official pump prices for Special 95, Super 98 and Diesel. Use today's date to work out which month is next. Do all the research, data gathering and modelling yourself, then deliver it as a single interactive HTML dashboard I can keep and share.

1. DATA
- Official monthly UAE Fuel Price Committee prices (AED per litre) for Special 95, Super 98 and Diesel, from January 2022 to the current month. Use official or reputable news sources, and cite them.
- Daily market prices for the same period: Brent crude, US Gulf Coast gasoline (RBOB) and diesel (ULSD/heating oil) futures. Convert them to AED per litre using the fixed 3.6725 AED/USD peg.
- UAE prices track the previous month's world product prices, so build monthly averages lagged by one month.

2. MODELS
Build these six models and combine them into an ensemble weighted by backtest performance:
- Static pass-through regression on refined product prices
- Dynamic pass-through with lagged terms
- Crude price + refining margin decomposition
- Asymmetric error correction model, to test whether rises pass through faster than cuts ("rockets and feathers")
- Kalman filter with time-varying pass-through
- Simple crude-only regression as a baseline
Use refined product prices as the main driver. They usually explain UAE prices better than crude alone.

3. VALIDATION
- Run a rolling-origin backtest from January 2024 onward. At each step, train only on data available at the time.
- Report direction accuracy (did it call up or down correctly), mean absolute error in AED, and each model's ensemble weight.
- Build prediction intervals with conformal prediction, calibrated on the backtest errors.

4. NOWCAST
The current month isn't over, so the input average isn't final. Simulate the remaining trading days from recent futures volatility, then feed the simulated averages through the ensemble. Report the probability of a price cut versus a rise for each fuel.

5. DASHBOARD
- A forecast card for each fuel: point forecast, a range bar, the current price, and the % chance of a cut or rise
- Attribution showing how much of each change comes from crude cost, refining margin and catch-up
- A what-if slider for "what if oil moves X% before month-end"
- A market signals table with the latest prices and their trend
- Charts of historical official prices against market prices and margins
- A transparency section with backtest accuracy, model weights, key coefficients, methodology and sources
- Clean, mobile-friendly design that works in light and dark mode

6. HONESTY
State clearly that this is an estimate, not financial advice. The committee doesn't publish its formula. Flag any data gaps or assumptions you made.

When you finish, give me a short summary of next month's forecast for each fuel, plus the backtest direction accuracy.