Straight answers,
including the uncomfortable ones.

Yes — demo paper ($100K simulated, Alpaca) to learn on, or live real money once you fund your account and connect your broker (Alpaca by default). All fills use real market prices, and every order requires your explicit approval through your own broker keys. We never take custody of funds.

Only you — every time, by design. The engine's authority is scan → gate → card → watch → report. An order exists only after you hit Approve on a card, and it is executed through your own broker account with your API keys (Alpaca by default). No approval → nothing happens. There is no auto-trade mode and none is planned.

Entries run 30–60 days to expiry (35 is the sweet spot). Typical close happens at the 50%-profit mark around 2–3 weeks. That's roughly 8 cycles per year per slot. Positions are never held 90 days "and hope" — theta accelerates hard under 30 DTE.

Three endings: bought back at +50%, expires worthless (keep full premium), or assigned — in which case the wheel takes over: you own 100 shares at the effective basis and sell covered calls against them until called away. Every ending is planned before entry.

The broker catalog lists 14,464 active names, but only 6,180 have real, tradeable option chains — the rest are warrants, units, preferreds, and phantom "optionable" flags with empty chains. Scanning ghosts would just manufacture fake candidates.

It's the real market tape. The free data tier is 15 minutes delayed — because every entry is a limit order and triggers are report-only, that delay is harmless; quotes are pulled fresh at order time. Charts render self-hosted via Apache ECharts.

Deliberately. This is the cashflow playbook: collect premium with a statistical edge and defined endings, instead of buying options where "many losers plus rare monster winners" is the expected shape. Consistency compounds; lottery tickets don't.

Honest math, two slots on a $100K account: roughly $100–120/month (~1.3%/yr). Six slots: ~$300/month. The theoretical ceiling for premium-selling at reasonable sizing runs around 8%/yr — more capital scales it, nothing miraculous happens. It's a consistency machine, not a moonshot. Anyone promising you more is selling you something.

Symbol + live price, the exact contract (strike/expiry/DTE), premium, breakeven, cushion %, the Greeks table for that contract, pre-set exits (+50% target, 2× credit stop), and the fresh earnings-date check. See the anatomy example.

Rule #1 has no pinch exception — that's the point. Any order that moves money requires an explicit approval, every time, including "harmless" wire tests and exploratory closes. The ledger proves the rule works: after a backwards-fill lesson (−$134, fully disclosed), two gates were added and the same mistake can't repeat.

Windows 64-bit and Linux x64 native clients, plus installable mobile apps on iOS 14+ / Android 9+, plus the web dashboard itself. On macOS, run the web dashboard in the browser for now.

During the paper phase, nothing — it's a private research demo. The commercial product (approval-gated cards for subscribers) is deliberately parked until there's a live track record, planned as a flat subscription under $100/yr. No performance fees — your P&L stays yours.