First-15-min opening-range break + confluence (gamma · flow · sector). Alerts fire at grade ≥ B; auto-mirror at grade A.
How this bot works. Watches ~46 liquid names for an opening gap
(|open vs prev close| ≥ desk floor) that breaks its first-15-minute
opening range in the gap direction. Each candidate is graded
A/B/C on confluence: GEX regime alignment, options-flow share in the
gap direction, and sector ETF sentiment alignment — desk-set thresholds, one signal stream for
every user. Alerts fire at grade ≥ B; auto-mirror fires only on grade A,
before the trigger cutoff (default 11:00 ET).
Execution is per user: an ATM contract is picked against your own broker quotes, sized by
your risk-%-per-trade against your equity, scaled by the market-wide gamma regime
(full size in positive gamma, damped in negative), and managed by the trail engine with your
own hard stop / trailing stop / profit target / wall-touch exits.
Watched
—
last scan: —
ORB Break (A)
—
all confirmations
ORB Break (B)
—
≥ 50% applicable
Armed
—
OR locked, awaiting break
Gappers
—
top: —
SortFilter
Loading…
⚙
Settings
runtime knobs · admin
Gap detection
Open vs prev close — default 1.0%. Sub-gap tickers ignored.
No new triggers after this time · default 11:00
Discord + dashboard fire at this grade or higher · default B · also acts as gate before auto-mirror
Confirmations
% of TD Pro bullish premium today needed to confirm long · default 0.60
Same for short side · default 0.60
Sector sentiment ≥ this confirms long · default +25
Sector sentiment ≤ this confirms short · default −25
Auto-mirror
When ON, alerts at min-grade or above auto-place a Tradier order
Further restriction on top of alert threshold · must be ≥ alert grade · default A⚠ Auto-mirror grade is looser than alert grade — only alerts that fire reach auto-mirror, so the effective threshold is your alert grade.
Trail engine
Close at −X% premium · default 33%
Close when premium drops X% from high · default 30%
Take profit at +X% premium · default 100%
Fire wall-touch this far before the wall · default 0.25%
🧪
Grade Lab
does grade predict the edge? · adminshadow · paper
Live gap only enters on grade A, so we never see how B/C actually perform. The
shadow tracker records every triggered setup at all grades (paper, no real
entry) and measures the underlying's directional move from trigger to that day's close — the signal's
raw edge by grade. Build a sample before changing the floor (aim for ≥40 per grade).
Grade
N
Win rate
Avg fwd return
Best
Worst
Loading…
💵
Real fills by grade
closed hunts · option P&L
Grade
N
Win rate
Avg P&L %
Total $
Loading…
Only grades that entered appear here — C never auto-mirrors, so it can't show up in real fills. That's what the shadow table above is for.
🤖
AI Review
what's working · what's not · tuning recommendationson demand · Anthropic API call
Window:model: configured via ANTHROPIC_MODEL
Click Generate Review to send the last window's gap-continuation alerts and hunt outcomes
to Claude for a quantitative critique with concrete tuning recommendations.
Requires ANTHROPIC_API_KEY set on the server.