/daily-review

description: Run a daily reflection on recent trades and behavioral patterns
argument-hint: "[date or 'today']"

Daily Review

Trigger an AI-powered reflection on recent trading activity. Analyzes trades, identifies behavioral patterns, updates affective state, and produces actionable insights.

Workflow

Step 1: Determine Review Period

If a date is provided, review that day's trades. Default: review all trades since last reflection.

Step 2: Gather Data

Load the `trading-memory` skill context, then:

1. Use `get_strategy_performance` to pull recent trade stats 2. Use `get_behavioral_analysis` to check behavioral drift 3. Use `get_agent_state` to read current affective state 4. Use `recall_memories` with recent context to surface relevant patterns

Step 3: Analyze Patterns

The reflection should cover:

**Trade Execution Quality**

• Did entries match the strategy rules? Or were they impulsive?
• Were stop losses honored? Any manual overrides?
• Position sizing: consistent with risk rules or erratic?

**Behavioral Drift Detection**

• Disposition effect: cutting winners, holding losers?
• Revenge trading: increased size after losses?
• Overtrading: more trades than the strategy signals justify?
• Session discipline: trading outside designated sessions?

**Strategy Performance**

• Which strategies fired today?
• Win/loss breakdown per strategy
• Any strategy consistently underperforming?

Step 4: Update Affective State

Based on the review, the affective state should be recalibrated:

• Confidence: up after good execution, down after poor discipline
• Risk appetite: reduce after drawdown, normalize after recovery
• Streak awareness: flag tilt risk after consecutive losses

Step 5: Produce Report

Structure:

## Daily Review — [Date]

### Summary
- Trades today: N (W wins, L losses)
- P&L: $XXX
- Best: [trade details]
- Worst: [trade details]

### Behavioral Check
- Disposition ratio: X.X (target < 1.0)
- Hold time balance: [OK / Winners cut short / Losers held too long]
- Position sizing: [Consistent / Erratic]

### Insights
1. [Specific, data-backed observation]
2. [Specific, data-backed observation]

### Tomorrow's Focus
- [One concrete action item based on today's data]

Important Notes

• Requires trades in the database. If no recent trades, report "no activity" instead of generating fluff.
• LLM reflection requires `ANTHROPIC_API_KEY`. Without it, uses rule-based analysis (still useful, less nuanced).
• Be honest. If performance is bad, say so. No sugar-coating.

Example

User: /daily-review today