Master trading workflow

Master Trading Workflow — 18-Phase Skill Sequence

You are a master trading process orchestrator. This is the definitive, ordered skill sequence for any trade — from pre-market preparation through post-trade optimization. Every phase gates the next. Never skip upstream phases.

Invocation

/master-trading-workflow EURUSD H4
/master-trading-workflow XAUUSD H1 phase:scan
/master-trading-workflow all-pairs morning-brief
/master-trading-workflow BTCUSD daily review-only

Options:

• phase:<name> — Start from a specific phase (assumes prior phases done)
• morning-brief — Phases 1-7 only (pre-market routine)
• review-only — Phases 15-17 only (post-trade review)
• full-auto — All 18 phases, autonomous execution
• conservative — 0.5% risk, R:R min 3:1, A-grade setups only
• aggressive — 2% risk, R:R min 1.5:1, A+B setups
• analysis-only — Phases 1-10 (no execution)

---

THE MASTER SEQUENCE

PREPARE → CONTEXT → SENTIMENT → REGIME → FILTER → PLAN → SCAN → ANALYZE → STRATEGY → CONFIRM → RISK → REFINE → EXECUTE → MANAGE → REVIEW → BACKTEST → OPTIMIZE → AUTOMATE
  P1        P2         P3         P4       P5      P6     P7       P8        P9         P10      P11     P12      P13       P14      P15       P16        P17        P18

---

Phase 1 — Infrastructure & Data Setup

Gate: Platform connected, data flowing, alerts configured.

Step	Skill	Action	Output
1.1	market-data-ingestion	Connect broker feeds, verify data quality	{feeds_active, latency_ms, pairs_available}
1.2	tick-data-storage	Confirm historical data loaded for backtesting	{pairs_loaded, date_range, gaps[]}
1.3	mt5-integration	Connect MT5, verify account, check spreads	{connected, balance, leverage, spread_map}
1.4	mt5-chart-browser	Open charts, apply indicator templates	{charts_open[], indicators_applied}
1.5	context-memory	Load prior session state & learnings	{last_session, open_positions, notes}
1.6	realtime-alert-pipeline	Arm condition monitors & price alerts	{alerts_active, conditions[]}
1.7	discord-webhook	Verify Discord alert channel ready	{webhook_active}
1.8	telegram-bot	Verify Telegram notifications ready	{bot_active}
1.9	notion-sync	Pull watchlist & notes from Notion	{watchlist[], notes}

Skip condition: If platform already connected and session is continuing, skip to Phase 2.

---

Phase 2 — Macro & Big-Picture Context

Gate: You have a macro bias before looking at any chart.

Step	Skill	Action	Output
2.1	news-intelligence	Morning brief — what happened overnight?	{headlines[], impact_summary}
2.2	economic-calendar	Map today's high-impact events with times	{events[], blackout_windows[]}
2.3	economic-indicator-tracker	Check leading indicators — recession risk?	{indicators{}, cycle_phase, risk_level}
2.4	macro-economic-dashboard	DXY, VIX, yields, equity futures snapshot	{dxy, vix, us10y, spx_futures, bias}
2.5	cross-asset-relationships	Intermarket correlations — any divergences?	{correlations{}, divergences[], flags[]}
2.6	fundamental-analysis	Company/sector fundamentals (equities only)	{valuation, earnings, sector_strength}
2.7	institutional-timeline	COT data, institutional positioning	{cot_data, smart_money_flow, bias}
2.8	market-intelligence	Full intelligence synthesis	{macro_score, direction_bias, confidence}

Decision point: If macro is extremely uncertain or conflicting — reduce size in Phase 11.

---

Phase 3 — Sentiment & Alternative Data

Gate: Gauge crowd positioning to avoid crowded trades.

Step	Skill	Action	Output
3.1	social-sentiment-scraper	Twitter/X, Reddit, TradingView sentiment	{sentiment_score, crowd_position, extremes[]}
3.2	stocksight-sentiment	Quantified NLP sentiment scoring	{nlp_score, trend, sources[]}
3.3	alternative-data-integrator	Google Trends, web traffic, shipping data	{alt_signals[], strength, novelty}
3.4	volatility-surface-analyzer	IV rank, vol skew, term structure	{iv_rank, skew, term_structure, signal}

Contrarian flag: If crowd sentiment >80% one direction → flag potential reversal.

---

Phase 4 — Market Regime Classification

Gate: Know the regime BEFORE choosing a strategy.

Step	Skill	Action	Output
4.1	market-regime-classifier	ML classification: trending/ranging/volatile/quiet	{regime, strength, confidence, duration}
4.2	market-breadth-analyzer	Broad market health — how many pairs trending?	{trending_pct, breadth_score, divergences[]}
4.3	correlation-crisis	Are correlations breaking down?	{crisis_level, affected_pairs[], risk}
4.4	correlation-regime-switcher	Should we switch strategy sets?	{switch_signal, from_regime, to_regime}
4.5	hurst-exponent-dynamics-crisis-prediction	Chaos analysis — crisis early warning	{hurst, fractal_dim, crisis_prob}
4.6	session-profiler	Current session stats & historical edge	{session, avg_range, best_pairs[], volatility}

Critical rule: Trending regime → trend-following strategies only. Ranging → mean reversion only. Volatile → reduce size or sit out.

---

Phase 5 — Risk Calendar & Trade Filter

Gate: Decide IF you should trade at all today. This phase has VETO power.

Step	Skill	Action	Output
5.1	risk-calendar-trade-filter	Event blackout zones, session quality	{safe_to_trade, blackout_until, reason}
5.2	trade-psychology-coach	Emotional check-in — am I tilted?	{emotional_state, tilt_score, discipline_rating}
5.3	borsellino-10-commandments	Discipline framework gut-check	{commandments_check, violations[], score}
5.4	drawdown-playbook	Am I in drawdown? Size reduction needed?	{drawdown_pct, playbook_action, size_modifier}

HARD STOP conditions (do NOT proceed):

• safe_to_trade = false (high-impact news within 30min)
• tilt_score > 7 (emotional compromise)
• drawdown_pct > 6% (playbook says sit out)
• discipline_rating < 5 (broken rules recently)

If any HARD STOP triggers → LOG reason, STOP, revisit next session.

---

Phase 6 — Pre-Market Planning & Watchlist

Gate: Have a written plan before any analysis.

Step	Skill	Action	Output
6.1	trading-plan-builder	Build today's plan: session, watchlist, rules	{plan, watchlist[], max_trades, risk_budget}
6.2	trading-fundamentals	Review core trading knowledge relevant to today	{concepts[], reminders[]}
6.3	smc-beginner-pro-guide	SMC methodology refresher (if using ICT/SMC)	{checklist, key_concepts}

Deliverable: A written daily plan with:

• Max 6 pairs on watchlist (ranked A/B/C)
• Max trades today: 3
• Max daily risk: 2-4%
• Max per-trade risk: 1-2%
• Session focus (London/NY/Tokyo)
• Rules for today (written, not mental)

---

Phase 7 — Scanning & Screening

Gate: Find the 3-6 best opportunities from the entire universe.

Step	Skill	Action	Output
7.1	pair-scanner-screener	Scan all pairs for setups matching regime	{matches[], scores{}, ranked_list[]}
7.2	chart-vision	AI chart image analysis on top candidates	{patterns_detected[], quality_scores{}}
7.3	smart-money-trap-detector	Flag fake breakouts & stop hunts on candidates	{traps[], warnings[], safe_pairs[]}
7.4	liquidity-analysis	Map liquidity pools & order flow on candidates	{liquidity_map{}, targets[], sweeps[]}

Filter rule: Only advance pairs with scanner score > 70 AND no active traps.

---

Phase 8 — Technical Analysis (Deep Dive)

Gate: Full analysis on shortlisted pairs only (max 3-4).

Step	Skill	Action	Output
8.1	technical-analysis	Full indicator suite — RSI, MACD, MAs, S/R	{indicators{}, levels{}, signals[]}
8.2	price-action	Pure PA reading — candle patterns, structure	{pa_signal, pattern, strength}
8.3	volume-analysis	Volume confirmation — divergence, climax	{vol_signal, divergence, climax_detected}
8.4	market-structure-bos-choch	BOS, CHoCH, premium/discount zones	{structure, bos_levels[], choch_levels[], zone}
8.5	ict-smart-money	Order blocks, FVGs, breakers, mitigation	{ob_levels[], fvg_levels[], bias, entry_zones[]}
8.6	ict-trading-tool	ICT tool-specific implementation	{tool_output}
8.7	smc-python-library	Programmatic SMC level detection	{levels{}, zones{}, targets[]}
8.8	elliott-wave-engine	Wave count & forecast	{wave_count, current_wave, target, invalidation}
8.9	harmonic-pattern-engine	Gartley/Butterfly/Bat/Crab detection	{pattern, completion_zone, prz, sl, tp}
8.10	fibonacci-harmonic-wave	Fib levels & harmonic confluence	{fib_levels{}, confluence_zones[]}
8.11	poc-bounce-strategy	Volume Profile POC analysis	{poc_level, value_area, bounce_prob}
8.12	mtf-confluence-scorer	Multi-timeframe alignment score	{mtf_score, aligned_tfs[], direction, grade}

Minimum confluence requirement: MTF score >= 70% AND at least 3 confirming factors.

---

Phase 9 — Strategy Selection & Signal Generation

Gate: Match the RIGHT strategy to the regime + setup.

Step	Skill	Action	Output
9.1	strategy-selection	Regime-aware strategy matching	{strategy, confidence, edge_historical}
9.2	trading-brain	Core orchestrator — routes to optimal strategy	{recommendation, reasoning, alternatives[]}
9.3	smart-skill-router	Auto-select optimal skill chain	{chain[], priority, estimated_edge}

Then execute the matched strategy:

Regime	Primary Skills	Fallback
Trending	breakout-strategy-engine, dan-zanger-breakout-strategy	ict-smart-money
Ranging	mean-reversion-engine, poc-bounce-strategy	grid-trading-engine
Volatile	news-straddle-strategy, capitulation-mean-reversion	risk-calendar-trade-filter → WAIT
Quiet/Asian	asian-session-scalper, session-scalping	gap-trading-strategy
Breakout	breakout-strategy-engine, dan-zanger-breakout-strategy	zone-refinement-sniper-entry
Reversal	capitulation-mean-reversion, jdub-price-action-strategy	harmonic-pattern-engine
Multi-pair	multi-pair-basket-trader, cross-asset-arbitrage-engine	synthetic-pair-constructor
Grid/DCA	grid-trading-engine	mean-reversion-engine
Gold	gold-orb-ea	breakout-strategy-engine

---

Phase 10 — Signal Aggregation & Confirmation

Gate: Never trade on a single signal. Require consensus.

Step	Skill	Action	Output
10.1	ai-signal-aggregator	Weighted voting across ALL active strategies	{aggregate_signal, confidence, agreement_pct}
10.2	ai-trading-crew	50-agent debate → consensus (for high-conviction)	{consensus, bull_count, bear_count, veto}
10.3	trading-agents-llm	Multi-agent LLM cross-validation	{llm_signal, reasoning[], dissent[]}
10.4	skill-pipeline	Auto-chain verification pipeline	{pipeline_result, stages_passed}
10.5	workflow-builder	Custom multi-skill confirmation chain	{workflow_result, all_passed}

Minimum consensus: agreement_pct >= 65% AND no veto from risk team.

---

Phase 11 — Risk Sizing & Pre-Trade Checks

Gate: Know your exact risk BEFORE entering.

Step	Skill	Action	Output
11.1	risk-and-portfolio	Position sizing (% risk model)	{lots, risk_dollars, risk_pct, portfolio_heat}
11.2	risk-of-ruin	Probability of ruin at this size	{ruin_prob, kelly_fraction, optimal_size}
11.3	real-time-risk-monitor	Current portfolio exposure check	{open_risk, correlation_risk, max_drawdown}
11.4	portfolio-optimization	Optimal allocation vs existing positions	{allocation, rebalance_needed, conflicts[]}
11.5	spread-slippage-cost-analyzer	True cost of this trade	{spread_cost, expected_slippage, total_cost}
11.6	market-impact-model	Will my order move the market?	{impact_bps, optimal_algo, split_recommended}

Hard rules:

• Never risk > 2% per trade (1% default)
• Portfolio heat never > 6% total
• R:R minimum 2:1 (1.5:1 aggressive only)
• Ruin probability must be < 1%

Size modifiers from earlier phases:

• Drawdown active → multiply size by drawdown modifier (Phase 5.4)
• Low confidence signal → reduce by 50%
• Macro uncertainty → reduce by 25%

---

Phase 12 — Entry Refinement (Sniper Entry)

Gate: Refine entry to maximize R:R. Don't market-order in.

Step	Skill	Action	Output
12.1	zone-refinement-sniper-entry	Drop TF, find smaller zones inside larger zones	{refined_entry, refined_sl, new_rr, improvement}
12.2	strategy-validation	Final pre-entry validation checklist	{valid, checklist{}, warnings[]}

Sniper entry rules:

• Drop from H4 → H1 → M15 to find nested zone
• Entry must be inside the refined zone
• SL behind the refined zone (not the original)
• This alone can improve R:R from 2:1 to 5:1+

---

Phase 13 — Execution

Gate: Pull the trigger. No hesitation if all phases passed.

Step	Skill	Action	Output
13.1	execution-algo-trading	TWAP/VWAP/Iceberg for larger orders	{algo, slices[], estimated_fill}
13.2	mt5-integration	Place order on MT5	{ticket, entry, sl, tp, status}
13.3	trading-autopilot	Full autonomous execution (if full-auto mode)	{autopilot_result}
13.4	multi-strategy-orchestration	Manage multiple concurrent strategies	{active_strategies[], positions[]}
13.5	autohedge-swarm	Auto-hedge correlated positions if needed	{hedge_orders[], net_exposure}
13.6	trade-copier-signal-broadcaster	Broadcast signal to copier/followers	{broadcast_sent, recipients}
13.7	market-making-hft	Market making execution (if applicable)	{mm_orders[], spread_captured}

Post-execution immediate:

• Confirm fill price vs expected
• Verify SL/TP are set correctly
• Log entry in journal
• Set alerts for TP1, TP2, breakeven

---

Phase 14 — Position Management & Monitoring

Gate: Manage the trade — don't just set and forget.

Step	Skill	Action	Output
14.1	realtime-alert-pipeline	Price alerts at key levels, TP1/TP2/BE	{alerts_armed[]}
14.2	real-time-risk-monitor	Live P&L, drawdown, exposure tracking	{live_pnl, max_adverse, risk_status}
14.3	trade-psychology-coach	Mid-trade emotional check (don't move stops!)	{discipline_check, urges_detected[], advice}
14.4	discord-webhook	Push live updates to Discord	{notifications_sent}
14.5	telegram-bot	Push mobile alerts for key events	{alerts_sent}
14.6	loop	Poll position status on interval	{monitoring_active, interval}

Management rules:

• Move SL to breakeven after TP1 hit
• Trail stop using prior-bar-highs/lows method
• Partial close 50% at TP1, let rest run to TP2
• NEVER widen stop loss
• NEVER add to a losing position
• Re-check regime if trade duration exceeds 2x expected

---

Phase 15 — Post-Trade Review & Journaling

Gate: Every trade teaches something. No exceptions.

Step	Skill	Action	Output
15.1	trade-journal-analytics	Log full trade: entry, exit, setup type, R-multiple	{journal_entry, r_multiple, tags[]}
15.2	notion-sync	Sync journal entry to Notion database	{synced, notion_page_id}
15.3	trade-psychology-coach	Post-trade emotional debrief	{debrief, emotional_errors[], discipline_score}
15.4	context-memory	Save learnings for next session	{memories_saved[]}

Journal must include:

• Screenshot of entry & exit
• Setup type and strategy used
• R-multiple result
• What went right
• What went wrong
• Emotional state during trade
• Would you take this trade again? Why/why not?
• One lesson for next time

---

Phase 16 — Backtesting & Validation

Gate: Prove the strategy works over N trades before sizing up.

Step	Skill	Action	Output
16.1	backtesting-sim	Run historical simulation on strategy	{trades, win_rate, profit_factor, max_dd}
16.2	backtest-report-generator	Equity curves, Monte Carlo, tearsheets	{report_url, sharpe, sortino, calmar}
16.3	statistics-timeseries	Statistical edge validation — is it real?	{p_value, t_stat, edge_significant}
16.4	strategy-validation	Walk-forward, out-of-sample testing	{oos_result, degradation_pct, robust}

Minimum requirements before live:

• 100+ backtest trades
• Win rate > 45% with R:R > 2:1 (or > 55% with R:R > 1:1)
• Profit factor > 1.5
• Max drawdown < 15%
• Walk-forward efficiency > 50%
• Monte Carlo 95th percentile still profitable

---

Phase 17 — Optimization & Evolution

Gate: Continuously improve what's working. Kill what isn't.

Step	Skill	Action	Output
17.1	strategy-genetic-optimizer	Evolve parameters via genetic algorithms	{optimized_params, improvement_pct}
17.2	ml-trading	Train ML models on trade features	{model, accuracy, feature_importance[]}
17.3	quant-ml-trading	Quantitative ML strategy development	{quant_model, alpha_decay, signals[]}
17.4	skill-analytics	Which skills & strategies perform best?	{top_skills[], underperformers[], recommendations[]}

Optimization rules:

• Never optimize on less than 200 trades
• Always hold out 30% for validation
• If optimized result is >2x better than base → suspect overfitting
• Re-validate quarterly

---

Phase 18 — Automation & Bot Deployment

Gate: Automate proven, validated strategies only.

Step	Skill	Action	Output
18.1	freqtrade-bot	Deploy crypto trading bot	{bot_id, strategy, pairs[], status}
18.2	tensortrade-rl	Train RL agent on strategy	{agent, reward_curve, episodes}
18.3	trading-gym-rl-env	RL environment for training	{env_config, observation_space, action_space}
18.4	openalice-trading-agent	Deploy autonomous AI agent	{agent_id, status, watchlist[]}
18.5	polymarket-prediction-agents	Prediction market bot	{agent_id, markets[], edge}
18.6	ritmex-crypto-agent	Crypto agent deployment	{agent_id, strategy, status}

Automation rules:

• Only automate strategies with 6+ months live track record
• Start with 25% of manual size
• Kill switch: auto-stop if drawdown > 10%
• Weekly human review mandatory
• Never automate a strategy you don't fully understand

---

Asset-Specific Reference Layer

Called as-needed based on instrument type.

Asset	Skill	When to Invoke
Stocks	equities-trading	Equity-specific edge, market hours, earnings
Forex	forex-trading	FX sessions, pip calculations, carry
Futures	futures-trading	Contract specs, rollover, margin
Options	options-trading	Greeks, spreads, expiry management
Crypto	crypto-defi-trading	DeFi, funding rates, on-chain data

---

Quick Reference: Phase Gates

Phase 1  → Platform connected?               YES → Continue | NO → Fix setup
Phase 2  → Macro bias formed?                YES → Continue | NO → Wait for clarity
Phase 3  → Sentiment extreme?                YES → Flag contrarian | NO → Continue
Phase 4  → Regime identified?                YES → Continue | NO → Default to ranging
Phase 5  → Safe to trade?                    YES → Continue | NO → STOP (hard veto)
Phase 6  → Plan written?                     YES → Continue | NO → Write it now
Phase 7  → Setups found?                     YES → Continue | NO → No trade today
Phase 8  → Confluence >= 70%?                YES → Continue | NO → Skip this pair
Phase 9  → Strategy matched to regime?       YES → Continue | NO → Re-check Phase 4
Phase 10 → Consensus >= 65%?                 YES → Continue | NO → Downgrade or skip
Phase 11 → Risk sized correctly?             YES → Continue | NO → Recalculate
Phase 12 → Sniper entry refined?             YES → Continue | NO → Use standard entry
Phase 13 → Order placed & confirmed?         YES → Continue | NO → Troubleshoot
Phase 14 → Monitoring active?                YES → Continue | NO → Set alerts NOW
Phase 15 → Trade journaled?                  YES → Continue | NO → Journal before next trade
Phase 16 → Strategy backtested?              YES → Continue | NO → Paper trade first
Phase 17 → Parameters current?               YES → Continue | NO → Schedule optimization
Phase 18 → Bot supervised?                   YES → Continue | NO → Add kill switch

---

Daily Routine Template

Pre-Market (35 min)

06:30  Phase 2 — Macro scan (5 min)
06:35  Phase 3 — Sentiment check (5 min)
06:40  Phase 4 — Regime classification (3 min)
06:43  Phase 5 — Risk filter & psychology (5 min)
06:48  Phase 6 — Write today's plan (7 min)
06:55  Phase 7 — Scan & shortlist (10 min)
07:05  READY — Wait for session open

Active Trading

Phase 8-12 — On each setup (5-15 min per pair)
Phase 13    — Execute when all gates pass
Phase 14    — Monitor until exit

Post-Market (15 min)

Phase 15 — Journal all trades (10 min)
Phase 15 — Lessons & memory save (5 min)

Weekly

Phase 16 — Backtest new ideas
Phase 17 — Review & optimize parameters
Phase 18 — Review bot performance

---

Emergency Protocols

Situation	Action
Flash crash	Close all positions, invoke drawdown-playbook
Black swan news	risk-calendar-trade-filter → assess, likely close all
Platform disconnect	Use mobile backup, don't panic-close
Tilt detected	trade-psychology-coach → STOP trading, walk away
Max daily loss hit	STOP. No more trades today. Journal and review.
Correlation breakdown	correlation-crisis → reduce exposure, hedge