stock-analysis-workflow

name	stock-analysis-workflow
description	股票/ETF 投资分析工作流：本地 DuckDB 定量分析 + 可靠来源定性分析 + 标准报告。Use when: 分析、投资价值、估值、基本面分析、ETF估值、股票研究。
argument-hint	股票/ETF代码或名称 + 分析目标
user-invocable	true

Stock Analysis Workflow Skill

This skill is for investment decision analysis. It complements stock-quant-analysis, which focuses on quantitative strategy backtesting.

Scope Boundary

• Use this skill for: valuation, fundamentals, industry context, bull/bear thesis, qualitative logic, and written reports.
• Use stock-quant-analysis for: Alpha158 backtests, technical signal backtests, Chanlun evaluation, and strategy optimization.
• Use analysis-report after finishing to store the final report in analysis/<stock_code>/.

Standard Workflow

1. Clarify the target stock/ETF if ambiguous.
2. Check local historical reports in analysis/<stock_code>/ and _index.json.
3. Prefer local DuckDB data for quantitative analysis.
4. Write a Python script under ./tmp/<session_id>_* for calculations.
5. Use external sources only for information that cannot be quantified locally.
6. For external information, prefer reliable primary sources and cite them.
7. Synthesize results into a structured report.
8. Save the report with analysis-report.

Report Structure

Reports must include:

1. Introduction
2. Data and Methodology
3. Quantitative Results
4. Qualitative Analysis
5. Conclusion and Risks

Rules

• Do not skip local DuckDB checks when data may exist.
• Do not use unverified web claims as evidence.
• Do not write final reports outside analysis/ except temporary drafts in tmp/.
• Do not make unsupported price predictions.