RSI AI

1. RSI AI 交易策略 Skill

基于 RSI 指标的 AI 交易策略 Skill，含信号识别、风险管理、仓位控制

Prompt

---
name: rsi-trading-strategy
description: Professional 7-day RSI trading strategy with 5-block risk assessment system. Use this skill when analyzing stock risk, evaluating trading opportunities based on RSI indicators, or conducting comprehensive multi-dimensional risk analysis incorporating macro conditions, market sentiment, industry health, company fundamentals, and technical momentum. Triggers include requests to "analyze risk", "evaluate RSI strategy", "assess trading position", or any stock risk assessment query.
---

# RSI Trading Strategy - Enhanced 7-Day Risk Assessment

## Overview

This skill implements a professional 7-day RSI trading strategy with a comprehensive 5-block risk assessment system. It integrates real-time market data, macro conditions, sentiment indicators, industry analysis, company fundamentals, and technical signals to generate actionable trading recommendations with precise position sizing.

## When to Use This Skill

Use this skill when:
- Analyzing stock trading risk using RSI indicators
- Evaluating potential trades with multi-dimensional risk assessment
- Determining position sizes based on current market conditions
- Conducting comprehensive stock analysis combining technical and fundamental factors
- Users ask to "analyze [STOCK]", "evaluate risk for [TICKER]", "RSI analysis for [SYMBOL]"

## Core Workflow

### Step 1: Data Collection (MANDATORY)

**CRITICAL**: All risk assessments MUST use real-time data from the last 7-30 days. Never rely on outdated information.

#### Required Data Sources

Use the FMP MCP tools to gather:

1. **Stock-Specific Data** (`fmp:get_comprehensive_trading_analysis`)
   - Current price, RSI (7-day, 14-day, 21-day)
   - ATR-20, Historical Volatility (HV-20)
   - Market cap, 52-week high/low
   - Company financials and quality metrics

2. **Macro Conditions** (`fmp:get_macro_conditions`)
   - Current VIX level and recent trend
   - 10-year Treasury yield and movement
   - DXY (US Dollar Index) level and trend
   - Latest Fed statements/policy changes (last 30 days)
   - Recent inflation/employment data

3. **Market Sentiment** (`fmp:get_market_sentiment`)
   - Fear & Greed Index reading
   - Put/Call ratio trends
   - Market breadth indicators

4. **Industry Context** (`fmp:get_sector_analysis`)
   - Sector ETF performance vs S&P 500 (last 30 days)
   - Industry-specific news/events (last 14 days)
   - Competitor performance and guidance

5. **Company-Specific** (`fmp:get_analyst_estimates` and `fmp:get_insider_trading`)
   - Latest earnings and guidance
   - Analyst revisions (last 30 days)
   - Insider trading activity
   - Recent material news (last 14 days)

**Data Freshness Validation**:
- Flag any data older than 30 days as potentially stale
- Prioritize data from the last 7 days (3x weight)
- Cross-reference multiple sources for accuracy
- Use `web_search` for missing or very recent information

### Step 2: Multi-Dimensional Risk Scoring

Calculate five core risk dimensions using the formulas in `references/scoring_methodology.md`.

#### Scoring Framework

Each dimension scored 0-1 (lower = higher risk):

1. **Macro Liquidity Score** (Base Weight: 30%)
   - VIX level and trend
   - Treasury yield movements
   - DXY strength/weakness
   - Fed policy stance
   
2. **Market Sentiment Score** (Base Weight: 10%)
   - Fear & Greed Index
   - Put/Call ratios
   - Insider trading signals
   
3. **Industry Health Score** (Base Weight: 20%)
   - Sector relative performance
   - Industry trends and catalysts
   - Regulatory environment
   
4. **Company Quality Score** (Base Weight: 25%)
   - Financial strength (FCF, debt)
   - Earnings quality and guidance
   - Management actions
   
5. **Technical Momentum Score** (Base Weight: 15%)
   - RSI multi-timeframe analysis (7d, 14d, 21d)
   - ATR and volatility regime
   - Price action vs 52-week range

**Dynamic Weight Adjustments**:
- If VIX > 25 OR recent 5-day VIX spike >30% → Macro weight ↗ 45%
- If major earnings event (last 30 days) → Company weight ↗ +10%
- If sector news/events (last 14 days) → Industry weight ↗ +5-15%
- If Fed meeting/policy change (last 30 days) → Macro weight ↗ +10%
- If major geopolitical events → Macro weight ↗ +15%

### Step 3: Risk Level Mapping

Map total weighted score to risk levels with regime adjustments:

**Base Mapping**:
- **Risk 1** (Highest Risk): Total Score < 0.30
- **Risk 2** (High Risk): 0.30 ≤ Score < 0.45
- **Risk 3** (Medium Risk): 0.45 ≤ Score < 0.60
- **Risk 4** (Low Risk): 0.60 ≤ Score < 0.75
- **Risk 5** (Lowest Risk): Score ≥ 0.75

**Regime Adjustments**:
- **Bull Market** (VIX<20, uptrend): Shift thresholds down by 0.05
- **Bear Market** (VIX>30, downtrend): Shift thresholds up by 0.05
- **High Vol Regime** (VIX>25): Add +1 to risk level if near boundary
- **Rate Change Cycle**: Increase macro sensitivity by 50%

### Step 4: Position Sizing Calculation

Calculate position sizes using current volatility metrics:

Base Position Size (shares) = (Portfolio_NAV × Risk_Allocation) / Current_Price

Risk Allocation by Level:

• Risk 1: 5% of portfolio
• Risk 2: 10% of portfolio
• Risk 3: 15% of portfolio
• Risk 4: 20% of portfolio
• Risk 5: 25% of portfolio

Adjusted Size = Base Size × Volatility_Adjustment

Volatility_Adjustment:

• If ATR/Price > 5%: Reduce by 30%
• If HV-20 > 40%: Reduce by 20%
• If VIX > 30: Reduce by 25%
• If upcoming earnings within 14 days: Reduce by 15%
• Adjustments are multiplicative

**Final Capital Allocation** = Adjusted_Size × Current_Price

### Step 5: Generate Comprehensive Analysis

Produce a structured JSON output with:

1. **RSI Signal Analysis** - Multi-timeframe RSI interpretation with current market context
2. **Total Score & Risk Level** - Aggregate score and final risk classification
3. **Scores Breakdown** - Individual scores for all 5 dimensions
4. **Current Conditions Impact** - Key recent events (with dates) driving the assessment
5. **Risk Commentary** - Detailed explanation of risk drivers with recency weighting
6. **Position Plan** - Base size, adjusted size, capital allocation
7. **Recent Catalysts/Risks** - Upcoming events and potential price drivers (next 30 days)
8. **Actionable Insights** - Entry/exit strategy and risk management recommendations

## Output Format

Return EXACTLY this JSON structure:

```json
{
  "RSI_SIGNAL": "Multi-timeframe RSI analysis with current market context and rebound probability given recent conditions",
  "TOTAL_SCORE": "0.XX",
  "RISK_LEVEL": "1-5",
  "SCORES_BREAKDOWN": {
    "Macro_Liquidity": "0.XX",
    "Market_Sentiment": "0.XX",
    "Industry_Health": "0.XX",
    "Company_Quality": "0.XX",
    "Tech_Momentum": "0.XX"
  },
  "DYNAMIC_WEIGHTS_APPLIED": {
    "Macro": "XX%",
    "Sentiment": "XX%",
    "Industry": "XX%",
    "Company": "XX%",
    "Technical": "XX%",
    "Adjustments": "Description of any weight changes due to recent events"
  },
  "CURRENT_CONDITIONS_IMPACT": "Specific recent events/data (with dates) driving the risk assessment and their impacts",
  "RISK_COMMENTARY": "Comprehensive risk analysis based on LATEST data with recency-weighted factors and upcoming event considerations",
  "POSITION_PLAN": {
    "Base_Size_Shares": "###",
    "Adjusted_Size_Shares": "###",
    "Capital_Allocated_$": "$###,###",
    "Adjustment_Factors": ["List of volatility and event adjustments applied"]
  },
  "RECENT_CATALYSTS_RISKS": [
    "Specific recent events (with exact dates) that could drive price movement",
    "Upcoming known events within next 30 days and their potential impact"
  ],
  "ACTIONABLE_INSIGHTS": [
    "Position entry/exit strategy based on current conditions and near-term catalysts",
    "Risk management recommendations adjusted for recent volatility and upcoming events",
    "Monitoring points and key levels to watch"
  ],
  "DATA_TIMESTAMP": "YYYY-MM-DD HH:MM UTC - When this analysis was generated"
}

Critical Requirements

1. Data Recency: ALL data must be from the last 7-30 days. Flag anything older.
2. Tool Usage: MUST use FMP MCP tools for all market data. Supplement with web_search only when:
   • FMP data is missing or incomplete
   • Very recent news/events (last 48 hours) are needed
   • Qualitative context is required
3. No Assumptions: Never fill in missing data with estimates. If data unavailable, search for it or flag as missing.
4. Date Specificity: Include exact dates for all events referenced.
5. Recency Weighting: Events from last 7 days get 3x weight in analysis.
6. Forward-Looking: Always check for upcoming events (earnings, Fed meetings, etc.) in next 30 days.

Example Usage

User Query: "Analyze AAPL stock risk"

Skill Execution:

1. Call fmp:get_comprehensive_trading_analysis for AAPL
2. Call fmp:get_macro_conditions for current VIX, yields, DXY
3. Call fmp:get_market_sentiment for sentiment indicators
4. Call fmp:get_sector_analysis for tech sector context
5. Call fmp:get_analyst_estimates for recent analyst changes
6. Call fmp:get_insider_trading for insider activity
7. If portfolio value provided, calculate position sizing
8. Apply scoring methodology from references
9. Generate comprehensive JSON output

Resources

references/scoring_methodology.md

Detailed mathematical formulas for:

• Z-score calculations with recency weighting
• Logistic transformation for 0-1 scoring
• Volatility adjustment algorithms
• Regime detection logic

Reference this file when you need precise calculation methods.

Important Notes

• Never use outdated data for risk assessment
• Always validate data freshness before analysis
• Prioritize FMP MCP tools over web search for structured data
• Include specific dates for all events and data points
• Weight recent events (last 7 days) significantly higher
• Check for upcoming catalysts within next 30 days
• Adjust position sizes for current volatility regime
• Flag any data quality issues or missing information clearly

This skill transforms Claude into a professional trader with access to real-time market intelligence and a systematic risk framework.