π₯THE PHOENIX AI AGENT TRADING PROTOCOLπ₯
I built a rulebook that lets an
#AI agent swing trade a small account β starting at $1,000 β with discipline most humans can't hold.
Here's how it works, in plain English:
π¦βπ₯ THE HUNT
The agent screens for strong stocks (RSI above 60) having a "red day" β leaders on sale, not laggards. For downside trades, it flips the logic: weak underlying stocks bouncing into resistance, traded through inverse ETFs.
π¦βπ₯ THE CONFIRMATION
No trade enters on prediction. Price has to prove it β a break back above the pullback high (or below the bounce low).
#Wyckoff structure and volume decide trade quality; the
$VIX decides trade size. The market confirms, or the agent sits in cash.
π¦βπ₯ THE ARMOR
β’ 2% standard risk per trade, 5% ceiling on rare A-grade setups
β’ Real stop-loss orders at the broker the moment of entry β never "mental stops"
β’ No averaging down. Ever. No revenge trades. No
#FOMO.
β’ No new trades into CPI, FOMC, or jobs reports
β’ Kill switch: β15% drawdown = everything closes, trading halts
π¦βπ₯ THE HARVEST
Partial profits banked at 1R and 2R until the trade's cost is back in the account β then a small "free runner" rides the trend with a trailing stop, risking zero capital.
π¦βπ₯ THE REBIRTH
Every trade β and every trade the rules BLOCKED β gets logged. Each month, the data prunes what loses and keeps what wins. The system literally rebuilds itself from its own ashes. That's the "
#Phoenix".
Slow. Mechanical. Boring on purpose. The goal isn't to predict the
#market β it's to survive long enough to compound.
Not financial advice. Trading involves real risk of loss, and leveraged ETFs amplify it. π¦βπ₯
(see agent prompt below)β¬οΈ
drive.google.com/file/d/1y9fβ¦