Chat with James
Daily Workflow
Scheduled Tasks
LIVE ALERTS
Confluence Bot proximity alerts
β
| Prompt | Schedule | Status | Next Run | Last Run | |
|---|---|---|---|---|---|
| Loading... | |||||
YouTube Channels
| Channel Name | Channel ID | Added | |
|---|---|---|---|
| Loading... | |||
Memory
| Content | Type | Salience | Date | |
|---|---|---|---|---|
| Loading... | ||||
Trading Dashboard
Loading...MORNING BRIEF
No digest yet β runs weekdays at 9am ET. You can also ask James to run it now via Telegram.
MACRO
Loading market data...
SECTORS SPDR ETFs β 1D / 1W / 1MO
Loading sector data...
SENTIMENT & BREADTH
CNN FEAR & GREED
β
Loading...
SECTOR BREADTH
Loading...
ADVANCE / DECLINE
Live A/D available with Schwab API
OPTIONS EXPIRATION
Loading...
PRE-MARKET MOVERS
Loading movers...
ECONOMIC CALENDAR This week Β· USD Β· High & Medium impact
Loading calendar...
EARNINGS CALENDAR S&P 100 Β· 3 days out Β· Finnhub
Loading earnings...
MARKET NEWS
Loading headlines...
STOCK WATCHLIST Top 20 by score Β· updates every 10 min
SCORE
SYMBOL
PRICE
CHG%
REL VOL
CATALYSTS
Loading watchlist...
OPENING REVERSAL SCANNER INACTIVE
Active 9:30–10:30am ET. Stocks down >2% with elevated volume are flagged as reversal candidates.
GEX / MARKET MAKER LEVELS Gamma wall Β· flip point Β· key strikes
MA CONFLUENCE CHECKER 21/55ema Β· 100ma Β· 200ma Β· 200ema β daily Β· weekly Β· 15min Β· hourly Β· 4h
LEVELS BOARD All known levels for a ticker β MAs, GEX, session H/L, VWAP, trendlines, S/R
Enter a ticker to load all levels.
All Bots Overview
BOT
MODE
REGIME
WR Β· W/L
TODAY
3D
7D
ALL-TIME
OPEN / UNR
LAST TRADE
BTC Bot
Autonomous BTC perpetual trader. Claude is the decider.
Every 4 hours (00, 04, 08, 12, 16, 20 UTC) Claude looks at BTC across 5m / 15m / 1H / 4H / 1D bars, a clustered S/R map (200/100/55/21 MAs across 1H/4H/1D/1W, daily and weekly pivots, swing H/L over 7/30/90/180d, auto-detected diagonal trendlines over 30/90/180d windows with β₯3 touches, round numbers), macro context (DXY, equities, VIX, gold, 10Y), BTC dominance, ETH/BTC ratio, and the last 12h of crypto news, plus the open position and the recent decision history. He returns one action β open long, open short, close, hold, trim, add β with a thesis, stop, target, size, and confidence. Paper trades fill at the agent's prices and a 1-minute monitor watches stop / target hits between 4H ticks.
PAPER ONLY
β ACTIVE
|
$10k notional Β· 1% risk Β· default 5x leverage Β· 4H decision cadence
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
WR:
β
|
Next tick:
β
SPY Bot LLM
Claude-as-decider Β· 1H cadence RTH Β· intraday only (EOD close 15:55 ET)
Loading...
CADENCE & CONTEXT
LLM ticks at 9:30 / 10:30 / 11:30 / 12:30 / 13:30 / 14:30 ET (manage-only past 14:00).
Each tick feeds Claude: SPY bars (30m/1h/daily) + macro (VIX/futures/yields) + sectors + breadth +
fear-greed + OpEx + GEX (call wall / put wall / flip point) + levels + open position + lessons.
Paper $10k Β· 1% risk Β· 1x leverage Β· single position Β· force-close at 15:55 ET.
Each Claude call ~10-15s Β· ~6 decisions/day Β· monitor every 60s between ticks for stop/target.
Each tick feeds Claude: SPY bars (30m/1h/daily) + macro (VIX/futures/yields) + sectors + breadth +
fear-greed + OpEx + GEX (call wall / put wall / flip point) + levels + open position + lessons.
Paper $10k Β· 1% risk Β· 1x leverage Β· single position Β· force-close at 15:55 ET.
Each Claude call ~10-15s Β· ~6 decisions/day Β· monitor every 60s between ticks for stop/target.
OPEN POSITION
none
RECENT DECISIONS (last 20)
| Time | Action | Conf | SPY | Regime | Thesis |
|---|
CLOSED TRADES (last 30)
| Entry | Exit | Dir | Entry $ | Exit $ | Stop | Target | Reason | Shares | P&L | % |
|---|
LESSONS
(none yet)
SPY Bot (Mech)
Mechanical paper trader Β· intraday 30m Β· BT-locked v1 config
Loading...
LOCKED v1 CONFIG
single-position Β· stop=0.4% Β· wall=0.3% Β· gateB3ByRegime='shorts-in-uptrend'
maxHold=13 30m-bars (1 trading day) Β· entry cutoff 14:00 ET Β· EOD close 15:30 ET
$10,000 paper account Β· 1% risk/trade Β· zero fees Β· long+short
maxHold=13 30m-bars (1 trading day) Β· entry cutoff 14:00 ET Β· EOD close 15:30 ET
$10,000 paper account Β· 1% risk/trade Β· zero fees Β· long+short
OPEN POSITION
none
RECENT DECISIONS (last 20)
| Time | Action | SPY | Regime | Playbook | Thesis |
|---|
CLOSED TRADES (last 30)
| Entry | Exit | Dir | Playbook | Entry $ | Exit $ | Stop | Target | Reason | Shares | Bars | P&L | % |
|---|
MA Cluster Bot
Multi-TF MA cluster Β· far-edge entries Β· BT in progressLoading...
OPEN POSITIONS
none
RECENT TRADES (last 50)
| Entry time | Ticker | Dir | Mode | Entry | Stop | Target | Exit | Reason | P&L |
|---|
BTC Bot (Mech)
Mechanical BTC paper trader β same rules as the BT engine.
Pure deterministic strategy validated at PF 1.60, +104%/yr, 15% DD over the last 365 days of BTC data. Each 4H tick (00, 04, 08, 12, 16, 20 UTC): classifies regime (4H+1D 55EMA), detects setups by priority β B3-default (breakout retest) β B1 (pullback in trend) β B2 (range fade), runs entry gates (C1 extended candle / C4 wall check / C4b zone-bias for B1), opens paper trade with stop at structural level + 0.8% buffer, target placed before the next opposing zone (with zone-2 fallback). Concurrent positions allowed up to 6% total open risk (3 trades Γ 2% each). Stop/target monitored every 1 minute between ticks. No LLM β same fixed rules every time.
PAPER ONLY
β ACTIVE
|
$10k notional Β· 2% risk Β· 6% cap Β· 5x lev Β· before-zone target with fallback
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
WR:
β
|
Equity:
β
BTC Bot (Mech 1H)
Mechanical BTC paper trader β 1H cadence, sweep-tuned + calendar-filtered (May 2026).
Same rule pipeline as the 4H Mech bot but ticks every UTC hour boundary and uses 1H-tuned knobs. Sweep id 6/7/8 on 8 months of data picked these defaults: stopBuffer 0.5%, maxHold 20 bars, wallPct 1.5%, minRR 1.2, waitForFavorableR 0.5, riskPct 2%, maxOpenRisk 8%, target before next zone. Gates: C4 wall check ON (essential); C1 / C4b / B3-express OFF (each hurts on 1H). 0.08% round-trip fee modeled. Paper-only; live execution wires in after real-paper validation.
Calendar filter: UTC Saturday skip.
Walk-forward + full-window head-to-head (May 2026) confirmed UTC-Sat bleeds while Asian Sunday-morning is the bot's best window. Filter adds +0.15 PF and +5 pts ROI on a $10k start vs no filter; PT-Sat variant was tested and underperforms (cuts into the good Sunday-morning hours). Manual β‘ Run Now ticks bypass the filter so you can test any day.
BT performance (Jun 2025βMay 2026, 10.5mo, $10k start): PF 1.88, +99.7%, 18.4% DD, 58% WR, 125 trades. 3y Monte Carlo median end equity from $5k: ~$245k. Edges considered: out-of-sample validated; not yet validated across multiple BTC regimes (data only extends to 2024-10-31; the test window was BTC -29.5% drawdown).
PAPER ONLY
β ACTIVE
|
$10k notional Β· 2% risk Β· 8% cap Β· 5x lev Β· 1H ticks Β· before-zone target Β· UTC-Sat skip Β· sweep-tuned May 2026
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
WR:
β
|
Equity:
β
BTC Pattern Trader
Chart pattern trader β detects on 4H, enters on 1H breakout, both textbook AND failure directions.
Detects double bottom, double top, head & shoulders (regular + inverted), and triangles (ascending / descending / symmetrical) on 4H bars. For each detected pattern the engine arms TWO conditional orders: the textbook breakout direction and the failure direction. Whichever fires first wins, the other is cancelled. Confluence scoring overlays clustered MA zones + daily/weekly pivots + projected D/W trendlines so a pattern with neckline at a key level ranks higher.
Status: BT-only mode. Backtester drives all R&D. Live execution wires in to ByDFi after the sweep finds a stable param set.
Sweep is CLI-driven: npx tsx scripts/sweep-btc-patterns.ts (uses the parallel worker pool tuned by parallel-tune.ts).
Sweep is CLI-driven: npx tsx scripts/sweep-btc-patterns.ts (uses the parallel worker pool tuned by parallel-tune.ts).
SELECTED RUN
id:
β
|
PF:
β
|
P&L:
β
|
Trades:
β
Β·
βW
Β·
βL
|
WR:
β
|
DD:
β
|
Detections:
β
LEG SPLIT
Textbook:
β
|
Failure:
β
LIVE STATUS
|
Last tick:
β
|
Mark:
β
|
Equity:
β
RISK / TRADE
In-sample DD β 22% for current risk Β· changes apply at next entry
Config:
run-770
SOL Pattern Trader
Chart pattern trader β detects on 4H, enters on 1H breakout, both textbook AND failure directions.
Detects double bottom, double top, head & shoulders (regular + inverted), and triangles (ascending / descending / symmetrical) on 4H bars. For each detected pattern the engine arms TWO conditional orders: the textbook breakout direction and the failure direction. Whichever fires first wins, the other is cancelled. Confluence scoring overlays clustered MA zones + daily/weekly pivots + projected D/W trendlines so a pattern with neckline at a key level ranks higher.
Status: BT-only mode. Backtester drives all R&D. Live execution wires in to ByDFi after the sweep finds a stable param set.
Sweep is CLI-driven: npx tsx scripts/sweep-sol-regime-filter.ts (uses the parallel worker pool tuned by parallel-tune.ts).
Sweep is CLI-driven: npx tsx scripts/sweep-sol-regime-filter.ts (uses the parallel worker pool tuned by parallel-tune.ts).
SELECTED RUN
id:
β
|
PF:
β
|
P&L:
β
|
Trades:
β
Β·
βW
Β·
βL
|
WR:
β
|
DD:
β
|
Detections:
β
LEG SPLIT
Textbook:
β
|
Failure:
β
LIVE STATUS
|
Last tick:
β
|
Mark:
β
|
Equity:
β
RISK / TRADE
In-sample DD β 22% for current risk Β· changes apply at next entry
Config:
run-770
Co-Trader
paste or drop a TradingView screenshot β train my readTrain me to read your charts.
Pick BTC or SOL above. Paste a TradingView screenshot (Ctrl+V) or drag one in. I'll parse the lines I see (horizontals + trendlines), describe them in plain English, and lay them out as a table you can correct. When you lock the interpretation, the corrections are saved and feed into my conventions. History is per-asset; conventions are shared.
Drop a screenshot here, paste (Ctrl+V), or click to browse
PNG or JPG Β· 10MB max
SOL Bot
Autonomous SOL perpetual trader. Claude is the decider.
Every 4 hours (00, 04, 08, 12, 16, 20 UTC) Claude looks at SOL across 5m / 15m / 1H / 4H / 1D bars, a clustered S/R map (200/100/55/21 MAs across 1H/4H/1D/1W, daily and weekly pivots, swing H/L over 7/30/90/180d, auto-detected diagonal trendlines over 30/90/180d windows with β₯3 touches, round numbers), macro context (DXY, equities, VIX, gold, 10Y), BTC dominance, SOL/BTC ratio, and the last 12h of crypto news, plus the open position and the recent decision history. He returns one action β open long, open short, close, hold, trim, add β with a thesis, stop, target, size, and confidence. Paper trades fill at the agent's prices and a 1-minute monitor watches stop / target hits between 4H ticks.
PAPER ONLY
β ACTIVE
|
$10k notional Β· 1% risk Β· default 5x leverage Β· 4H decision cadence
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
WR:
β
|
Next tick:
β
SOL Bot (Mech)
Mechanical SOL paper trader β same rules as the BT engine.
Pure deterministic strategy validated at PF 1.60, +104%/yr, 15% DD over the last 365 days of SOL data. Each 4H tick (00, 04, 08, 12, 16, 20 UTC): classifies regime (4H+1D 55EMA), detects setups by priority β B3-default (breakout retest) β B1 (pullback in trend) β B2 (range fade), runs entry gates (C1 extended candle / C4 wall check / C4b zone-bias for B1), opens paper trade with stop at structural level + 0.8% buffer, target placed before the next opposing zone (with zone-2 fallback). Concurrent positions allowed up to 6% total open risk (3 trades Γ 2% each). Stop/target monitored every 1 minute between ticks. No LLM β same fixed rules every time.
PAPER ONLY
β ACTIVE
|
$10k notional Β· 2% risk Β· 6% cap Β· 5x lev Β· before-zone target with fallback
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
WR:
β
|
Equity:
β
BTC Zones
Manually defined BTCUSDT price zones. Polls every 30 seconds and enters paper or live trades via ByDFi when price signals a reversal inside a zone. Supports both-direction and fixed-direction zones.
β Set a price zone, time window, TP, and SL β James watches for a reversal signal inside it
β Position size is 5% of account per trade Β· One trade per zone, then it's done
β Approach direction locks on first candle that touches the zone: price from below = Long, price from above = Short Β· Once locked it never resets
β Fixed direction: only fires if approach matches Β· Both: fires in whichever direction price actually approached from
β Signal = any 15-min candle that closes in the reversal direction while still inside the zone (Long: green close, Short: red close)
β Can be candle 1 or candle 10 β no limit, just must close inside zone before window ends
β Entry at market on signal candle close Β· TP and SL placed immediately after
β Live execution via ByDFi Β· 5% position size per trade
Loading account...
NEW ZONE
ZONES
No zones yet.
Confluence Bot BT
MA zone reversal Β· strategy backtest
Checking bar coverage...
FROM
TO
Run:
Select a run above, or configure dates and run a new backtest. Download bar data first if the coverage bar shows gaps.
Sector Timeline
90d past + 30d AI forecast Β· Relative strength vs SPYRegime read
Hit Regenerate Forecast to build the timeline.
SECTOR
NAME
NOW
Ξ 4W
PHASE
REASONING
No data yet.
βΆ
Macro data fed to the model
Trend Bot
Swing trend continuation Β· daily cadence Β· paper + liveβΌ Entry criteria β what the bot looks for
GATES (all must pass)
β Trend Β· 20d HH+HL vs prior 20d, close > 50ma
β Channel Β· pivots fit a rising line (RΒ² β₯ 0.6, β₯3 pivot lows)
β Pullback magnet Β· today's low within 1Γ ATR of channel support
β 21ema touch Β· low touched 21ema in last 5 bars
β Reversal Β· close > open Β· close > prior close Β· close > prior bar's high
β Volume Β· today's vol β₯ 0.8Γ 20d avg
β ATR Β· ATR(14) / price β₯ 1%
β Cooldown Β· β₯3 bars since prior trade in this ticker
EXECUTION
β Buy-stop @ signal_high Γ 1.001
β€· Skip if open gaps > 2% past trigger
β€· If high never breaks trigger β no trade
EXIT
‡ Stop · swing_low à 0.999 (5-bar low)
‴ 2R partial · sell 50%, stop to breakeven
β³ Trail rest Β· close-below-21ema (after partial)
β± Time stop Β· 30 trading days max
SIZING
$250 fixed risk per trade
shares = 250 / |entry β stop|
Mirror everything for shorts: 20d LL+LH, sell-stop below signal low, stop above swing high.
PAPER
LIVE
Variant: hhhl Β· Risk $250 Β· 2R partial Β· 21ema trail Β· 30d max hold
OPEN POSITIONS
| Entry | Ticker | D | Mode | Entry Px | Stop | 2R | Shares | Partial | Days Held | Variant |
|---|
RECENT SETUPS
| Trigger | Ticker | D | Sig Close | Trigger Px | Stop | 2R | Shares | ADR% | VolΓ | Ch RΒ² | Ch Dist (ATR) | SPY Slope | Status |
|---|
TRADE LOG
| Entry | Ticker | D | Mode | Entry Px | Stop | 2R | Exit | Exit Px | Reason | Days | R | P&L | SPY Slope |
|---|
SPY Bot BT
Mechanical-surrogate intraday backtest Β· 30m cadence Β· EOD close 15:30 ET
Checking bar coverage...
PRESET
Run:
First Red Day BT
FRD/FGD reversal backtest on stored ~150 GEX_BT_WATCHLIST tickers
β
Threshold:
Backtest the First Red Day strategy with stored historical data. Universe is restricted to GEX_BT_WATCHLIST (~150 large/mid-cap S&P 500 tickers β small-cap runners are not in our bar data so trade counts are lower than production). AI signal weight is set to 0 in BT (no historical Claude assessments) β confluence threshold is reduced to 6 to compensate. News gate is bypassed in BT (no historical Finviz news).
BT Strategy Recap
- Streak filter: 3-8 consecutive green/red days, β₯+20%/-15% return Β· price β₯$3 Β· ADR β₯3%
- Confluence score (max 12, threshold 6): vwap_side(2) + avwap_side(3) + pdl_pdh_break(2) + vwap_reject(3) + bar_quality(2). AI signal removed.
- Stop = HOD/LOD Β± 0.1% buffer Β· Target = 30% ADR Β· $250 risk/trade
- No new entries after 11:30 ET Β· EOD market close at 3:55 PM ET (no overnight)
PERFORMANCE BY CONFLUENCE SCORE
| Score | Trades | Wins | Win% | Avg P&L | Total P&L |
|---|---|---|---|---|---|
| No run loaded. | |||||
BY DIRECTION
| Dir | Trades | Wins | Win% | Avg Win | Avg Loss | P&L | PF |
|---|---|---|---|---|---|---|---|
| No run loaded. | |||||||
BY DOMINANT TRIGGER
| Trigger | Trades | Win% | Avg P&L | Total P&L |
|---|---|---|---|---|
| No run loaded. | ||||
TRADE LOG
| Date | Ticker | Dir | Streak | Score | Trigger | In | Out | Entry | Exit | Stop | Target | Sh | P&L | Result |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No run loaded. | ||||||||||||||
Swing Trend Bot BT
Trend continuation Β· daily bars Β· pullback to 21ema
Setup: trend filter passes · pullback to 21ema (last 5 bars) · today bullish/bearish reversal candle on volume > 0.8× 20d avg ·
enter next bar's open · stop = swing low/high × (1 ± 0.1%) · partial 50% at 2R, trail rest on 21ema close break · time stop 30 days · risk $250/trade
VARIANT
FROM
TO
Run:
Confluence BT
GEX + MA reversal backtestBacktests GEX wall levels as reversal zones. Snapshots levels at end of day and tracks whether price touched and reversed, building a historical performance record.
β Logs GEX levels (Flip Point, Gamma Wall, Breakdown, Key Levels) for a core watchlist each weekday at 4:35pm ET Β· Only key levels β₯20% of the Gamma Wall's GEX qualify
β Flags when a GEX level coincides with an MA/EMA within 0.25% β or stands alone Β· Regular hours only (9:30amβ4pm ET)
β Touch = price comes within 0.15% of the level on a 15-min bar
β Reversal = price moves 0.75% away from the touch Β· Label reflects actual behavior (Support = bounced up, Resistance = rejected down)
Loading stats...
Loading data...
Zanger Scanner
6 chart patterns Β· S&P 500Scans S&P 500 for six Dan Zanger breakout patterns: Cup & Handle, High Tight Flag, Ascending Triangle, Flat Base, Double Bottom, Trend Line Breakout. Scored by proximity to breakout level, volume ratio, and ADR.
Hit Scan to find Zanger pattern setups across the S&P 500.
Confluence Setup
Stocks near stacked MA zonesScans the S&P 500 for stocks within 1.5% of a stacked MA confluence zone. Results grouped by bias: clean longs, clean shorts, intraday-only. Run on demand.
Hit Scan to find stocks near confluence zones.
AMD BT
GEX + MA confluence reversal backtestTracks AMD confluence zone setups historically. Research prototype for the Confluence Bot strategy.
Loading...
Autotrader
Reversal setups at GEX wallsScans GEX (gamma exposure) levels across a watchlist as potential reversal zones. End-of-day snapshots feed the Confluence BT backtest.
Strategy: Reversal at GEX Walls
- Universe: Top 50 S&P 500 stocks by market cap (no ETFs)
- Trend filter: 4H 55EMA β longs only above, shorts only below
- Level filter: Gamma Wall (WL) only, within 1.5% of price Β· Flip Point (FL) shown on chart for reference but never used as entry β FL accelerates moves, doesn't reverse them
- Entry proximity: 5m 21EMA β no longs above it, no shorts below it (confirms price is stretched into the level)
- Entry trigger: Wick touches level (low for long, high for short) β subsequent candle closes back above (long) or below (short) the level Β· Cannot enter on the touch candle itself
- Invalidation: If price traded more than 10% of ADR through the level on the wrong side at any point today, skip the setup
- Stop: 10% of ADR from entry Β· Target: 25% of ADR from entry
- Size: 5% of account Β· Hard max 1% account risk per trade
- Not yet built: Volume filter Β· Opening reversal scanner Β· Order execution Β· Daily loss limit
Hit Scan to load setups. ~1 min for 50 stocks.
Swingbot Sectors
James learning to swing trade. SPY/QQQ tracked as macro context. XLK/XLE are the trade targets. Morning before the open, James writes predictions. EOD scores them. Friday synthesizes the week. You can veto any open prediction or override its close.
Loading...
Continuation Runners
End-of-day scan for stocks making large moves on elevated volume. AI rates each for continuation probability. Tracks 15-day price performance by signal type.
Hit Scan Now or wait for the daily run after market close.
Cont. Run. Swing
multi-day continuation, stop at day 0 low, trail 9 EMA dailyLong-only multi-day swing on Continuation Runners scanner output.
Trades a runner on day 1 after the strong day. After 9:35 ET, enters on the first 5m bar that closes bullish above VWAP and above day 0 HIGH (a confirmed breakout above the entire day-0 range, not just day-0 close). Skips the session if price has already wicked below day 0 low. Stop is day 0 low β 2% (wide, intentional β we're holding multi-day). Trails on the daily 9 EMA: if today's projected close drops below the 9 EMA, exits at the 3:55 PM ET bell. Day-5 BE stop-raise on any session β₯5 held with close > entry. Target overlays (60d swing high, next $5 round, 200d MA) stored for analytics only. $250 risk per trade.
PAPER TRADING
β ENABLED
|
day 1 / day 2 windows Β· $250 risk/trade Β· stop at day 0 low β 0.1%
REAL MONEY
β DISABLED
|
Schwab LIMIT entry + bracket OCO (holds stop overnight); 9 EMA trail fires at 3:55 ET via market order
ALL-TIME
P&L:
β
|
Closed:
β
Β·
β
W
Β·
β
L
|
Win rate:
β
Β·
Avg:
β
|
Open:
β
ACTIVE WINDOW β β
Premarket Runner
premarket news-catalyst runner, hard stop at entry-day low, trail 9 EMA dailyCatalyst-first premarket runner β long-only.
The 8:00am ET scan finds premarket gappers (β₯10%) on real news and Claude rates the catalyst (the primary gate β float/chart are context). After the open, enters LONG when a 5m bar closes above the premarket high (bullish, above VWAP), with an anti-chase guard; no new entries after 10:30 ET. Hard stop at the entry-day low β 2%. Holds multi-day, trailing the daily 9 EMA β the trail only ARMS once price closes above the 9 EMA (or after 5 days), so a close 9 EMA at entry can't stop you out early. Day-5 breakeven. No partials, no EOD force-close. Exits are yours to manage β held-position news check + a manual Close button. $250 risk per trade.
PAPER TRADING
β ENABLED
|
$250 risk/trade Β· stop at entry-day low β 2% Β· 9 EMA daily trail
REAL MONEY
β DISABLED
|
Schwab LIMIT entry + bracket OCO (holds stop overnight); 9 EMA trail fires at 3:55 ET via market order
ALL-TIME
P&L:
β
|
Closed:
β
Β·
β
W
Β·
β
L
|
Win rate:
β
Β·
Avg:
β
|
Open:
β
TODAY'S RUNNERS β β
Confluence Bot
Autonomous intraday engineAutonomous intraday MA-confluence trader on S&P 500 stocks.
Zones form where 200/100/55/21 MAs across 1D / 1W / 4H / 1H / 15m cluster within 0.2% along with daily and weekly pivots and GEX levels. The 4H 55EMA filters direction (price above = longs only, below = shorts only). Entry triggers when a 5-minute bar wicks the zone and the next bar closes back on the correct side. Stop is 10% ADR, target 30% ADR, $250 risk per trade, max 2 trades per ticker per day, no entries after 10:15 AM ET. Modes: PAPER (always on, simulated), LIVE (Schwab, off by default β toggle per session).
PAPER TRADING
β ENABLED
|
$250 risk per trade
Β·
β
REAL MONEY
β DISABLED
|
Schwab:
β
avail
Β·
β
BP
Β·
β
portfolio
REGIME GATE
LOADING
|
SPY 20d ATR:
β
(Lβ₯βΒ·Sβ€β)
|
SPY
β
return:
β
(Lβ₯βΒ·Sβ€β)
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
Win rate:
β
Β·
Avg:
β
Autonomous intraday paper trading engine. Builds a zone watchlist at 9am, monitors every 5 minutes for touches and bounces, and enters/exits trades automatically. Sends Telegram proximity alerts when price approaches a zone.
Strategy: MA Confluence Zone Reversal
- Universe: S&P 500 stocks Β· price β₯$10 Β· avg vol β₯4M Β· ADR β₯1%
- Zones: Cluster of β₯2 levels within 0.2% of each other Β· also folds in daily & weekly pivots (dPivot/dR1/dS1/dR2/dS2, wPivot/wR1/wS1/wR2/wS2) and GEX levels (CW/PW/ZG) if available
- MAs included: Daily β 200ma, 200ema, 100ma, 55ema Β· Weekly β 200ma, 200ema, 55ema Β· 4H β 200ma, 200ema, 100ma, 55ema Β· 1H β 200ma, 200ema Β· 15min β 200ma, 200ema
- MAs excluded: 21ema on all timeframes Β· 100ma-weekly Β· 100ma/55ema on 1H and 15min
- Trend filter: 4H 55EMA β longs only above, shorts only below Β· zone must be on the correct side of price
- Violation check: If price has already closed more than 25% ADR through the zone, zone is marked done and skipped
- Entry trigger: 5-min bar wick touches zone (low β€ zone center for long, high β₯ for short) β next bar closes back on correct side Β· entry via limit order at bounce bar close Β±$0.01 (long +$0.01, short -$0.01) Β· EOD close is market Β· 21EMA(5m) proximity not required (differs from GEX Scanner)
- Stop: 10% ADR from entry Β· Target: 30% ADR from entry
- Size: $250 fixed risk per trade Β· Max 2 trades/day per ticker
- EOD: All open positions closed at last price at 3:55pm ET
- Proximity alert: Telegram fired when price comes within 0.2% of a zone center (before entry trigger)
TRADE LOG
| Date | Ticker | Dir | Bucket | In | Out | Entry | Exit | Target | Stop | Shares | P&L | Result | VIX | Sector% | SPY | AD | 4H EMA% | Zone | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Select a date above. | |||||||||||||||||||
SETUPS SEEN
All qualifying zones from morning scan
| Date | Ticker | Bias | Zone | Center | Dist% | Lvls |
|---|---|---|---|---|---|---|
| Select a date above. | ||||||
First Red Day Bot
Multi-day-streak reversal traderReversal trader for extended runners.
Shorts stocks that ran up 3β8 consecutive days (β₯+20% total) on the first red day; longs stocks that dropped 3β8 days (β₯-15%) on the first green day. Universe is the YF gainers/losers screeners plus the GEX watchlist, filtered by price β₯$3, avg volume β₯500K, market cap β₯$50M, ADR β₯3%. Claude scores each setup high_quality / parabolic_exhaustion / crowded_squeeze_risk / unclear (1β5). Pre-open re-verification, intraday state machine waits for VWAP rejection or PDL/PDH break, $250 risk per trade, max 1 trade per ticker, no new entries after 11:30 AM ET, EOD market close.
PAPER TRADING
β ENABLED
|
$250 risk per trade Β· 1 trade/ticker Β· cutoff 11:30 ET
REAL MONEY
β DISABLED
|
Schwab live trading via existing API config
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
Win rate:
β
Β·
Avg:
β
Identifies stocks that have run UP 3+ consecutive days (FRD short candidates) or DOWN 3+ consecutive days (FGD long candidates). Post-close scan finds setups, AI assesses each for reversal quality, intraday bot enters on PDL/PDH break with VWAP confirmation.
Strategy: First Red Day / First Green Day Reversal
- Holding period: Intraday only β every open position is market-closed at 3:55pm ET. Nothing held overnight.
- Universe: Yahoo screeners (most_actives, day_gainers/losers, small_cap_gainers/losers) + GEX_BT_WATCHLIST Β· price β₯$3 Β· avg vol β₯500K Β· mcap β₯$50M Β· ADR β₯3%
- Streak filter: 3-8 consecutive green days with total run β₯+20% (FRD short) Β· or 3-8 consecutive red days with total drop β₯-15% (FGD long). Uses only COMPLETED daily bars β today's still-developing close is ignored until 4:01pm ET.
- Scan timing: Post-close at 4:30pm ET weekdays. Pre-open re-verify at 8:00am ET with fresh headlines + AI re-assessment.
- AI assessment (Claude): Each setup classified high_quality / parabolic_exhaustion / crowded_squeeze_risk / unclear with 1-5 score (5 = highest conviction reversal)
- Intraday state machine: waiting β armed (price on correct side of prevClose AND VWAP) β entered (when weighted confluence score β₯ 8 + news gate passes)
- Confluence score (entry trigger, threshold 8/14): Six signals weighted and summed each tick:
- VWAP side (+2): price on correct side of session VWAP
- AVWAP side (+3): price is EXTENDED on the trade's side of anchored VWAP from open of streak day 0. For FRD shorts: entry > avwap (stretched above the up-streak's mean). For FGD longs: entry < avwap (capitulated below the down-streak's mean). Bigger extension = more stretched setup = better reversal candidate.
- PDL/PDH break (+2): 5m bar closes through prior day's low (short) / high (long)
- VWAP rejection (+3): 5m bar wicks to VWAP from wrong side + closes back on correct side + body confirms (bearish for shorts, bullish for longs)
- Bar quality (+2): bar body and close-position both align with direction
- AI signal (Β±2): high_quality +2, parabolic_exhaustion +1, unclear 0, crowded_squeeze_risk -2 (always counts)
- Pre-entry news gate: Right before placing the order, re-fetches current Finviz headlines and asks Claude whether any fresh CONTRADICTORY catalyst exists (positive news for shorts, negative for longs). If yes, trade is skipped and a Telegram β SKIP alert fires with the reason. Defaults to "go" on any infra failure.
- Stop: HOD + 0.1% buffer (FRD) or LOD - 0.1% (FGD) Β· Target: 30% ADR from entry Β· trigger_type column stores the dominant fired signal (vwap_reject / avwap_side / pdl_pdh_break / etc.)
- Size: $250 fixed risk per trade Β· Max 1 trade per ticker per day
- Cutoffs: No new entries after 11:30am ET Β· EOD market close at 3:55pm ET (no overnight holds)
- Tracking: 15 trading-day close tracking from trigger date for performance review by AI signal type
SETUPS
Streak Rider
Mid-cap continuation longContinuation long on mid-cap multi-day runs.
Longs mid-cap stocks ($2B-$10B mcap) that have run UP 3-7 consecutive days β₯+15%. Two intraday triggers: Gap-and-Go (open > prev close, first 3 bars hold above VWAP, enter on PDH break with bullish bar) or VWAP Pullback (price retraces to VWAP within first 90 min, enter on reclaim with bullish bar). Stop = intraday LOD - 0.1%. Target = +0.50 ADR$ (v1, will sweep in BT). $250 risk per trade Β· 1 trade/ticker Β· no entries after 11:30 AM ET Β· EOD market close 3:55 PM ET.
PAPER TRADING
β ENABLED
|
$250 risk per trade Β· 1 trade/ticker Β· cutoff 11:30 ET
REAL MONEY
β DISABLED
|
Schwab live trading via existing API config
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
Win rate:
β
Β·
Avg:
β
TODAY'S SETUPS
Trade Tickets
your trades, dictated in chat Β· bot monitors entries (paper)
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
Win rate:
β
Β·
Avg:
β
|
Triggers eval on completed 5m bars, RTH only Β· day tickets flat 15:55 ET
Ribbon Squeeze Bot
4HR ribbon squeeze β intraday morning breakout (long)Long the morning breakout out of a 4HR EMA-ribbon squeeze.
At 9:25 ET it arms tickers whose 4HR ribbon (14/21/34/55 EMAs) has compressed to a tight spread (β€0.6%), whose sector ETF is >EMA150 & <EMA8 (strong sector, short-term pullback), while SPY > SMA50. Intraday it enters the first 15m close above the ribbon top + above the 4HR 200-EMA (fresh cross) on rvol β₯ 3.2, between 9:45β11:00 ET. Exit: 0.5ΓADR hard stop + 15m-EMA12 trail armed @ 2 bars, force-flat 3:55 ET (89% intraday). $250 risk/trade, 1 trade/ticker. Paper-only β the backtest edge is recency-loaded, so this is forward validation.
PAPER TRADING
β ENABLED
|
$250 risk/trade Β· 1 trade/ticker Β· entries 9:45β11:00 ET Β· paper only
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
Win rate:
β
Β·
Avg:
β
Daytrade Prep
perp watchlist Β· ratings + trade ideasIntraday RS vs BTC Β· 5/15/30m
4h RS cross β red (fade watch)
FADE WATCH Β· recently running (fdp Β· pop β coil β break, newest first)
Daily Close Bot
PDC reclaim/reject traderTrades around prior day close (PDC) β the pivot that defines today's green vs red.
Pre-market scan reads each ticker's news, gap from PDC, sector ETF action, and macro context (SPY/VIX/futures/HYG/F&G/sector breadth) and asks Claude to direction it (LONG / SHORT / SKIP) with thesis and signal grade. Four setups depending on open side: reclaim_long (long bias, opened below PDC, 5m closes above), bounce_long (long bias, opened above PDC, pullback to PDC then bullish close), reject_short (short bias, opened above PDC, 5m closes below), retest_short (short bias, opened below PDC, rally to PDC then bearish close). Stop is PDC Β± 30% ADR (thesis dies if price moves through PDC); target is entry Β± 50% ADR (move AWAY from PDC). $250 risk, max 1 trade per ticker, no entries after 11:30 AM ET, EOD market close.
PAPER TRADING
β ENABLED
|
$250 risk per trade Β· 1 trade/ticker Β· cutoff 11:30 ET
REAL MONEY
β DISABLED
|
Schwab live trading via existing API config
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
Win rate:
β
Β·
Avg:
β
PDC (Prior Day Close) is the level that determines whether today is green or red. Pre-market scan reads news + gap + sector + macro and uses Claude to assign each ticker a directional bias (LONG / SHORT / SKIP). Intraday bot enters when price interacts with PDC in alignment with the bias β goal is to ride the move AWAY from PDC.
Strategy: Daily Close (PDC) Reclaim / Reject
- Holding period: Intraday only β every open position is market-closed at 3:55pm ET. Nothing held overnight.
- Universe: GEX_BT_WATCHLIST (~150 large/mid caps, ETFs excluded) Β· price β₯$5 Β· avg vol β₯1M Β· ADR β₯1.5%
- Pre-market scan (8:00 AM ET): fetches PDC/PDH/PDL + ADR + pre-market gap + sector ETF % + Finviz headlines + macro snapshot (SPY/VIX/futures/HYG/Fear&Greed/sector breadth)
- AI direction (Claude): Each ticker gets LONG / SHORT / SKIP with confidence 1-5 + catalyst + gap thesis. Direction must align with news + gap + sector + macro β don't fight the tape.
- Open classification: at first 5m bar, each ticker tagged
opened_above/opened_below/opened_atPDC (Β±0.1% deadband) - Setup A β Reclaim (long): LONG bias, opened below/at PDC, 5m bar closes above PDC β enter long
- Setup B β Bounce (long): LONG bias, opened above PDC, pulled back to PDC (bar low within 0.5% of PDC), 5m bar bullish (close > open) and closes above PDC β enter long
- Setup C β Reject (short): SHORT bias, opened above/at PDC, 5m bar closes below PDC β enter short
- Setup D β Retest (short): SHORT bias, opened below PDC, rallied to PDC (bar high within 0.5% of PDC), 5m bar bearish (close < open) and closes below PDC β enter short
- Stop: PDC Β± (0.30 Γ ADR$) β clean PDC-anchored stop. If thesis breaks (price moves 30% ADR through PDC the wrong way), trade is invalidated.
- Target: entry Β± (0.50 Γ ADR$) β meaningful move away from PDC. Roughly 1:1.5 R:R.
- Size: $250 fixed risk per trade Β· Max 1 trade per ticker per day
- Cutoffs: No new entries after 11:30am ET Β· EOD market close at 3:55pm ET
- Tracking: 15 trading-day close tracking from trigger date for performance review by AI signal type
SETUPS
Cont. Run. Intraday
long-only next-session continuationLong-only continuation trader for stocks that ran big the prior session on heavy volume.
Universe is the existing Continuation Runners scanner output (YF small-cap gainers + day gainers, filtered by % move + relative volume). Pre-market scan re-fetches multi-source headlines, refreshes the AI signal, and pulls pre-mkt gap. Intraday bot watches each tradeable ticker for one of four continuation patterns: ORB long (close above 9:30 ORB high, 9:35-10:00), VWAP reclaim (gapped down/flat, bar wicks below VWAP and closes back above), HOD break (after 10:00 ET, close above intraday HOD on volume), 9EMA pullback (wick to 9EMA, close above with bullish body). Stop = LOD - 0.1% (capped at 40% ADR$ below entry). Target = entry + 60% ADR$. $250 risk, max 1 trade/ticker, max 3 trades/day, no entries after 11:30 ET, EOD market close.
PAPER TRADING
β ENABLED
|
$250 risk/trade Β· max 1 trade/ticker Β· max 3 trades/day Β· cutoff 11:30 ET
REAL MONEY
β DISABLED
|
Schwab live trading (LIMIT + bracket OCO + EOD market close)
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
Win rate:
β
Β·
Avg:
β
SETUPS FOR β
LIVE RUNTIME
| Ticker | Cap | State | Last | VWAP | Open | ORB H | HOD | LOD | AI | Open Trade |
|---|---|---|---|---|---|---|---|---|---|---|
| No runtime data | ||||||||||
Volume Surge Bot
EOD daily-rvol scanner Β· continuation longs + gap-fail shorts Β· multi-day swing Β· scaffolding onlyEnd-of-day daily-rvol scanner. Continuation longs + gap-fail shorts. Multi-day swing.
Direction classifier turns each elevated-volume daily candle into a continuation_long, gap_fail_short, or skip. LLM Hook 1 vets at scan time (technical + news + fundamentals); Hook 2 is a pre-entry news gate; Hook 3 polls news on held positions at the open and again before the close. Mechanical exit is the base case; LLM and manual edits can override. Every scan / entry / stop / target / cutoff parameter is a sweep dimension at v1 β nothing is locked yet. Backtesting is the primary purpose of this build.
β Scaffolding only β no scan / no entries / no trades
DB tables, scheduler tasks, BOT_REGISTRY entry, API routes, Discord route, edit-trade plumbing all wired. Strategy logic lands in Task 4 (BT engine v0) and the live runtime is wired in Task 16 after sweeps lock a config.
PDH Group Bot
PDH rejection short basket β 8 tickersFailed-PDH short across VZ / CVX / MRK / PEP / AMD / MDT / GM / LLY.
Each ticker arms when a 5m bar wicks above prior-day high (PDH) with the prior bar's close still below PDH. Entry fires at the OPEN of the next bar. Quality gate: wick magnitude β€ 0.15 Γ ADR$. Stop at PDH + 5% ADR. Target at entry β 40% ADR. EOD market close 3:55 PM ET. $250 risk per trade, $50k notional cap. Validated PF 1.76, $30,958 over 24mo backtest.
PAPER
β ENABLED
|
$250 risk Β· 8 tickers Β· 9:35β15:00 ET entry window Β· EOD 15:55
LIVE
β DISABLED
|
Schwab integration deferred (paper-only v1)
ALL-TIME
P&L:
β
|
Trades:
β
Β·
Win rate:
β
Today's setups
Open positions
Recent trades
Mean Reversion Bot
velocity-event-armed intraday fadeVelocity-event-armed mean reversion. Fade institutional flow moves with no news.
A stock arms when its price has dropped or spiked by at least the price-band threshold inside a 5-min bar OR a rolling 40-min window (Dan's locked table). After arming, the bot waits for price to touch a confluence zone of β₯3 stacked MA/pivot levels; on close-back confirm, enters at the next 5m bar's open. Target = session_low + 80% Γ (session_high β session_low) for longs (mirror for shorts). Stop = zone outer edge (no buffer). $250 risk Β· max 1 long + 1 short per ticker per day Β· no entries after 11:30 ET Β· EOD safety close 3:55 ET. Sectors excluded: XLK, XLI, XLB. Pre-entry news gate via Claude on Finviz headlines (blocks on ANY material catalyst). Paper-only Phase 1.
PAPER TRADING
β ENABLED
|
$250 risk Β· target=0.80 retrace Β· stop=zone edge Β· cutoff 11:30 ET
REAL MONEY
β DISABLED
|
Schwab live trading stubbed pending paper-trade validation
ALL-TIME
P&L:
β
|
Trades:
β
News Trader
Theme Bot
News-driven intraday traderNews-driven theme trader.
Morning at 8:55 ET pulls overnight news and asks Claude to pick the 0β3 best themes of the day with tickers, setup types, and theses. Intraday (every 5 minutes from 9:30 to 15:55 ET) Claude evaluates setup-specific entry rules per ticker and manages open positions. EOD closes anything still open at last price. Paper-only β every setup type is tagged so we can study which kinds of news plays actually pay over time.
PAPER TRADING
β ENABLED
|
$250 risk per trade Β· max 2 trades/ticker Β· paper only
ALL-TIME
P&L:
β
|
Trades:
β
Β·
β
W
Β·
β
L
|
Win rate:
β
Each morning at 8:55 ET, Claude reads the news and picks the highest-conviction themes that are likely to move stocks today. For each affected ticker, it picks a setup type (gap_and_go / pullback / orb / fade / late_momentum). The bot then trades them intraday using simple, categorized entry rules so we can evaluate which setup types actually work over time.
TODAY'S THEMES
No themes loaded.
TRADE LOG
| Date | Ticker | Setup | Dir | Catalyst | In | Out | Entry | Exit | Stop | Target | Shares | P&L | Result |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Select a date above. | |||||||||||||
PERFORMANCE BY SETUP all-time
| Setup | Trades | Wins | Win% | P&L | Avg/trade |
|---|---|---|---|---|---|
| No trades yet. | |||||
PERFORMANCE BY CATALYST all-time
| Catalyst | Trades | Wins | Win% | P&L |
|---|---|---|---|---|
| No trades yet. | ||||
Dux Bots
Steven Dux's penny-stock long playbook
PMB CONTROL
Date:
View:
PRE-MARKET SETUPS
Loading setups...
Portfolio
Loadingβ¦
Trading Ops
Loading...
Wheel Scanner
Scans for wheel strategy candidates: elevated IV rank, liquid options chains, ranked for selling cash-secured puts.
| Ticker | Price | Strike | Expiry | Premium | Ann. Yield | IV30 | IVR | Earnings |
|---|---|---|---|---|---|---|---|---|
| Loading... | ||||||||
~0.25 delta front-month put Β· CSP yield = (premium Γ· strike) Γ (365 Γ· DTE) Β· Earnings within 21d excluded
Trendline Touch
Stocks at a fresh touch of their best-fit rising daily trendline. Scans all tickers with daily volume ≥ 1M. Per ticker, scores every (anchor_a, anchor_b) pivot-low pair by touches + post-confirmation touches minus violations, picks the highest-scoring line, and ranks the universe by distance from current close. Lines hard-broken more than 5 bars before today are filtered out.
| Ticker | Close | Line | Dist % | Low % | Slope/d | Tch | Post | Vio | Score | Age | Anchor A | Anchor B |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No scan yet. Click β» Scan Now (takes ~15-30s). | ||||||||||||
System Status
Agent
...
TTS Server
...
Active Tasks
...
Channels
...
Web Session
...