Rule number one isn't about trading.
It's about not trading.
Everything else is technique. These two rules outrank every signal, every hot read, every exception you'll want to make at 10:43 on a red morning. Real money once you fund it through your broker — demo paper $100K until you feel comfortable. The rules don't change between the two.
Run the desk in paper, or connect live funds and trade real money: both are supported, both are treated the same. Zero execution without explicit approval through your own broker (Alpaca by default) — no buys, no sells, no closes, no cancels, no "just testing the wire." Orders are placed with your broker keys, on your approval, in your account. One failed gate = the whole idea is done — there is no half-idea. Losses are findings, not tuition.
True 52-week window (intraday highs), Greeks on the exact contract, fresh earnings date, P/E, chart shape, IV, spread width — the full stack, every time. A card with a hole in its research is not a card.
What this engine sells — and never buys.
The signature position: sell puts on stocks you'd be happy to own at the strike. Premium up front, three peaceful endings.
Defined-risk version — sell one strike, buy protection below it. Both legs must quote cleanly before anything fires.
Assigned? Own the shares at an effective basis, sell covered calls against them, cycle out. The playbook was built for this.
Buying options produces many small losses and rare monster winners — a survivability problem. This engine collects premium; it doesn't buy lottery tickets.
Every put ends in exactly one
of three ways.
- +50% profit — buy back at half the credit sold, typically 2–3 weeks in. The disciplined exit.
- Expires worthless — keep the entire premium, position simply ends.
- Assigned — own 100 shares at effective basis and the wheel takes over with covered calls.
And the un-happiest case, handled by preset, not panic: stop at 2× the credit received. Small planned loss, logged, next candidate.
entry window 30–60 DTE (35 = sweet spot) typical close ~2–3 weeks at +50% cycles / year ~8 per slot never 90-day sits and hope theta reality accelerates under 30 DTE
Losses are printed, not buried.
A multi-leg order filled one leg while its opposite was already crossed — a backwards fill at −$0.14 debit on a trade designed for credit. The position was closed immediately and logged honestly. Result: gate 12 — both legs must quote within ~$0.25 before any multi-leg order fires, and the realized ledger shows every dollar exactly as it settled. The same audit produced the universal fresh-earnings gate after one missed report date.
Discipline is the product.
The dashboard prints wins and losses alike — the ledger is honest because you'll make better calls with honest numbers.