claude-code-stock-deep-research-agent
💡 Summary
A comprehensive AI-driven framework for conducting in-depth stock investment research.
🎯 Target Audience
🤖 AI Roast: “Powerful, but the setup might scare off the impatient.”
Risk: Medium. Review: shell/CLI command execution; outbound network access (SSRF, data egress). Run with least privilege and audit before enabling in production.
Claude Code Stock Deep Research Agent
Investment Research Edition - 专业股票投资尽调系统
⚖️ 免责声明
本研究报告不构成投资建议或推荐。所有投资存在风险,包括本金损失。
重要提示:
- 本报告仅供教育和信息用途
- 部分数据需要通过官方渠道验证
- 过往业绩不代表未来表现
- 投资决策前请自行进行尽职调查
- 建议咨询合格的财务顾问
🎓 研究框架
本研究基于 Claude Code Deep Research 系统:
- 方法论: 8阶段股票投资尽调框架
- 智能体: 28个并行研究智能体
- 工具: WebSearch、WebFetch、综合分析
- 质量: 多空平衡、明确风险、数据验证
Table of Contents
- Features
- Repo Structure
- Quick Start
- Stock Investment Research
- How It Works
- Customization
- Credits & Acknowledgements
- License
Features
This repository contains two specialized deep research frameworks for Claude Code:
1. 🎯 Stock Investment Research (股票投资尽调系统) ⭐ PRIMARY
An 8-phase investment due diligence framework for analyzing publicly traded companies, inspired by professional investment research methodologies.
Key Capabilities:
- 📊 Comprehensive Analysis: Business model, industry dynamics, financial quality, governance, valuation
- 🤖 Multi-Agent Research: ~28 parallel research agents working concurrently
- 📈 Investment Style Adaptation: Value, growth, turnaround, dividend investing
- 💰 Valuation Models: DCF, reverse DCF, relative valuation, scenario analysis
- 🛡️ Risk Assessment: Bear case, black swans, monitoring checklist
- ✅ Quality Assurance: Cross-validation (profit vs. cash flow, company vs. peers)
- 📝 Structured Output: 20-file standardized due diligence report
Research Coverage:
- A-shares (A股) - 中国大陆股市
- Hong Kong stocks (港股)
- US stocks (美股)
- Other global markets
Output: Signal rating (🟢🟢🟢 Strong Buy / 🟡🟡🟡 Hold / 🔴🔴 Avoid) based on fundamental analysis
2. 📚 General Deep Research (通用深度研究系统)
A flexible 7-phase framework for general research topics (business, technology, academic, etc.).
Repo Structure
| File/Folder | Purpose | |-------------|---------| | CLAUDE.md | Master instructions for Claude Code | | .claude/skills/stock-question-refiner/ | Stock research question refinement skill | | .claude/skills/stock-research-executor/ | 8-phase investment due diligence executor | | .claude/commands/stock-research.md | Main stock research command | | .claude/skills/citation-validator/ | Citation verification skill | | .claude/skills/got-controller/ | Graph of Thoughts controller | | .claude/skills/synthesizer/ | Findings synthesis skill | | STOCK_RESEARCH_IMPLEMENTATION_PLAN.md | Stock research system design document | | CLAUDE2.md | Graph of Thoughts implementation details | | PROJECT_UNDERSTANDING.md | Architecture deep dive | | IMPLEMENTATION_GUIDE.md | User guide |
Quick Start
Stock Research (股票投资尽调)
# Start Claude Code claude # Set model (optional, but recommended) /model opus # Execute stock research /stock-research [股票代码或公司名称] # Examples: /stock-research 600519 # 贵州茅台 (A-share) /stock-research AAPL # 苹果公司 (US) /stock-research 腾讯 00700.HK # 腾讯控股 (HK)
The system will:
- Ask about your investment style (价值/成长/困境/红利), holding period, risk tolerance
- Deploy ~28 parallel research agents across 7 phases
- Generate comprehensive due diligence report in
RESEARCH/STOCK_[ticker]_[company]/
Time: 2-4 hours for standard due diligence
Output Example
RESEARCH/STOCK_600519_Kweichow_Moutai/
├── 00_Executive_Summary.md # 🟡🟡🟡 Hold / Fairly Valued
├── 01_Business_Foundation.md # Products, revenue, customers
├── 02_Industry_Analysis.md # Industry cycle, competition
├── 03_Business_Breakdown.md # Profit drivers, economics
├── 04_Financial_Quality.md # Cash flow, margins, red flags
├── 05_Governance_Analysis.md # Ownership, management
├── 06_Market_Sentiment.md # Bull/bear cases
├── 07_Valuation_Moat.md # Moat rating, valuation
├── Financial_Data/ # Metrics, trends, peer comparison
├── Valuation/ # DCF, scenarios
├── Risk_Monitoring/ # Bear case, monitoring checklist
└── sources/ # Citations with quality ratings
Stock Investment Research
8-Phase Due Diligence Process
| Phase | Focus | Output | |-------|-------|--------| | 1. Business Foundation | 公司事实底座 | Products, revenue mix, customers, value chain, strategy | | 2. Industry Analysis | 行业周期分析 | Cycle stage, supply-demand, competition, policy impacts | | 3. Business Breakdown | 业务拆解 | Segments, profit engines, pricing power, economics | | 4. Financial Quality | 财务质量 | Metrics trends, cash flow vs. earnings, red flags, peers | | 5. Governance Analysis | 股权治理 | Ownership, management, capital allocation, ROIC | | 6. Market Sentiment | 市场分歧 | Bull case, bear case, key debates, verification nodes | | 7. Valuation & Moat | 估值护城河 | Moat rating (0-5), relative/absolute valuation, risks | | 8. Final Synthesis | 综合报告 | Signal rating, thesis, monitoring checklist |
Investment Style Adaptation
The system adapts research approach based on investment style:
| Style | Focus | Valuation Methods | Key Metrics | |-------|-------|-------------------|-------------| | Value (价值投资) | Intrinsic value, margin of safety | P/B, EV/EBITDA, DCF (conservative) | P/B, normalized earnings, FCF yield | | Growth (成长投资) | TAM, competitive positioning | PEG, DCF (aggressive), user models | Revenue growth, moat, TAM penetration | | Turnaround (困境反转) | Liquidity, solvency, catalysts | Liquidation value, option value | Debt ratios, cash runway, catalyst timeline | | Dividend (红利投资) | Dividend sustainability | DDM, FCF yield, yield comparison | Dividend yield, payout ratio, FCF coverage |
Quality Standards
Mandatory Cross-Validation:
- ✅ Profit vs. cash flow (OCF/NI ratio analysis)
- ✅ Company vs. peer comparison (key ratios, margins)
- ✅ Bear case analysis (3-5 key risks identified)
Citation Requirements:
- Every factual claim must include: Author, Date, Title, URL
- Source quality rated (A-E scale)
- A-grade sources: Annual reports, regulatory filings
- B-grade: Industry reports, analyst research
- C-grade: News articles, expert commentary
Signal Rating (based on fundamentals only):
- 🟢🟢🟢 Strong Buy: Significant margin of safety, strong moat, attractive valuation
- 🟡🟡🟡 Hold: Fairly valued or limited margin of safety
- 🔴🔴 Avoid: Overvalued, deteriorating fundamentals, excessive risks
Example Usage
User: /stock-research 贵州茅台 600519
Claude: 我可以帮您研究贵州茅台。让我先了解您的投资需求:
1. 投资风格:价值投资/成长投资/困境反转/红利投资?
2. 持有周期:短期(<6月)/中期(6-18月)/长期(1-3年以上)?
3. 关注重点:生意质量/财务质量/估值/公司治理 (选2-3个)
4. 风险偏好:保守/平衡/激进?
User: 价值投资,长期持有3年以上,关注:生意质量、公司治理、估值,风险保守
Claude: [Generating research plan...]
[Deploying 28 parallel research agents...]
[Analyzing 5-10 years of financial data...]
[Comparing with Wuliangye, Yanghe, Fenjiu...]
[Assessing competitive moat...]
[DCF valuation with 3 scenarios...]
[Identifying bear case risks...]
Output: RESEARCH/STOCK_600519_Kweichow_Moutai/
🟡🟡🟡 Hold / Fairly Valued
Investment Thesis:
茅台是中国领先的白酒品牌,拥有强大的品牌护城河和定价能力。然而当前估值(P/E 32x)安全边际有限。建议继续持有现有仓位,但新投资应等待10-15%回调至¥1,750-1,850区间。
Key Metrics:
- Market Cap: ¥2.8T
- P/E (TTM): 32x (above 5-year average)
- Gross Margin: 91.2%
- ROE: 31%
- Moat Rating: 5/5 (Very Strong)
Top 3 Reasons to Consider:
1. Unassailable brand moat with 800-year heritage
2. Exceptional margins (91% gross, 53% net)
3. Strong cash generation (OCF/NI > 1.0)
Top 3 Reasons to Avoid:
1. Full valuation (P/E 32x, limited margin of safety)
2. Regulatory risk (government scrutiny of luxury pricing)
3. Competitive intensification (Wuliangwa narrowing gap)
Monitoring Checklist:
✅ Strengthen: Price pulls back 10-15% to ¥1,750-1,850
❌ Exit: Price drops below ¥1,300 (-35%), net margin < 45%
[Full report: 20 files, 127 sources, 50+ pages]
How It Works
Stock Research Workflow
User: /stock-research [ticker]
↓
stock-question-refiner skill
- Asks: Investment style? Holding period? Focus areas? Risk tolerance?
↓
Structured Research Prompt (investment parameters, priorities, constraints)
↓
stock-research-executor skill
├─ Phase 1: Business Foundation (4 parallel agents)
├─ Phase 2: Industry Analysis (4 parallel agents)
├─ Phase 3: Business Breakdown (4 parallel agents)
├─ Phase 4: Financial Quality (4 parallel agents)
├─ Phase 5: Governance Analysis (4 parallel agents)
├─ Phase 6: Market Sentiment (4 parallel agents)
└─ Phase 7: Valuation & Moat (4 parallel agents)
↓
citation-validator skill
- Verifies all claims have citations
- Rates source quality (A-E)
↓
Comprehensive Investment Due Diligence Report
- Signal rating
- 8 phase reports
- Financial data tables
- Valuation analysis
- Risk monitoring checklist
Key Innovations
- Investment Style Adaptation: Research approach tailored to value, growth, turnaround, or dividend investing
- Parallel Multi-Agent Execution: ~28 agents working concurrently for efficiency
- Mandatory Cross-Validation: Profit vs. cash flow, company vs. peers, bear case analysis
- Structured Output: Standardized 20-file report format
- Quality Assurance: A-E source quality rating, citation verification
General Research Workflow (Secondary)
[ Question ] → [ stock-question-refiner ]
↓
[ Structured Prompt ]
↓
[ research-executor ]
├─ Planning (break into subtopics)
├─ Multi-Agent Research (parallel)
├─ Source Triangulation (A-E rating)
└─ Synthesis (combine findings)
↓
[ Citation Validation ]
↓
[ Research Report ]
Customization
Adapting Stock Research Parameters
The system automatical
Pros
- Comprehensive multi-agent research capabilities
- Structured output with standardized reports
- Adaptable to various investment styles
- Mandatory cross-validation for quality assurance
Cons
- Complex setup may deter casual users
- Requires understanding of investment concepts
- Time-consuming for comprehensive reports
- Dependent on data quality from external sources
Related Skills
pytorch
S“It's the Swiss Army knife of deep learning, but good luck figuring out which of the 47 installation methods is the one that won't break your system.”
agno
S“It promises to be the Kubernetes for agents, but let's see if developers have the patience to learn yet another orchestration layer.”
nuxt-skills
S“It's essentially a well-organized cheat sheet that turns your AI assistant into a Nuxt framework parrot.”
Disclaimer: This content is sourced from GitHub open source projects for display and rating purposes only.
Copyright belongs to the original author liangdabiao.
