Content
# Tool List
> "Price is what you pay, value is what you get." — Warren Buffett
>
> Redefining the depth and efficiency of investment research with AI.
**AI Berkshire** is a collection of investment research skills based on [Claude Code](https://claude.ai/code), systematically and structurally integrating the methodologies of four value investing masters: Warren Buffett, Charlie Munger, Duan Yongping, and Li Lu. Through AI Agents, it achieves professional-grade investment research.
One person + Claude = One investment research team.
---
## Real Track Record
> Not just theoretical discussions. This framework is backed by a real investment system verified with actual money.
### 2024 Full-Year Returns: +69.29%
<img src="assets/2024-returns.jpg" width="300" />
### 2025 Returns to Date: +66.38%
<img src="assets/2025-returns.jpg" width="300" />
### Comparison with Major Indices
| Indicator | 2024 Full-Year | 2025 to Date |
|------|----------|----------|
| **AI Berkshire Real Account** | **+69.29%** | **+66.38%** |
| Hang Seng Index | +17.67% | +27.77% |
| S&P 500 | +23.31% | +16.39% |
| CSI 300 | +14.68% | +17.66% |
| Nasdaq | +28.64% | +20.36% |
**2024 Excess Returns**: Outperformed S&P 500 by **46 percentage points**, outperformed Hang Seng Index by **52 percentage points**.
**2025 Excess Returns**: Outperformed S&P 500 by **50 percentage points**, outperformed Hang Seng Index by **39 percentage points**.
**Cumulative Real Account Returns over 2 Years Exceed RMB 1.46 Million**, significantly outperforming major global indices for two consecutive years.
> *Disclaimer: Historical returns do not represent future performance. Screenshots are from a real Futu Securities account.*
---
## Why Not Directly Ask AI?
You can directly ask Claude: "Help me analyze if Pinduoduo is worth buying." You will get a balanced analysis ending with "Investment involves risk, please make your own judgment."
**This kind of analysis looks okay but is not actionable.**
AI Berkshire solves not the "can it analyze" problem but the **analysis quality and decision-making discipline** issue. Here are the core differences:
### 1. Mandatory Conclusions, No Evasion
Directly asking AI gives you a balanced analysis. AI Berkshire outputs: **Pass/Not Pass/Gray Area**, with specific price ranges and tiered recommendations.
> Ordinary AI response: *"Pinduoduo has growth potential but also faces competitive pressure. Investors need to weigh..."*
>
> AI Berkshire Output:
> | Strategy | Suggestion | Price Range |
> |------|------|---------|
> | Aggressive | Build a 20% position at current price level | $95-105 |
> | Conservative | Wait for buyback policy clarification before building a position | $85-95 |
> | Cautious | Does not meet 10-year certainty criteria, observe | — |
>
> **Mirror Test**: If a 5-sentence summary cannot be completed, do not buy, with no exceptions.
### 2. Four Masters' Perspective Confrontation, Not Single Analysis
Not just "analyzing with Buffett's method." Four perspectives create **real conflicts and tensions**—
For example, with Pinduoduo:
- **Duan Yongping** (Business Model): Good business, C2M model hard to replicate → Score 3.7/5
- **Warren Buffett** (Financial Valuation): Cash-rich PE only 6.3x, money machine → Score 4.4/5
- **Charlie Munger** (Contrarian Thinking): Moat shallower than imagined, Douyin achieved 4 trillion GMV in 3 years → Score 3.5/5
- **Li Lu** (Long-term Certainty): Management culture has risks, uncertainty in 10 years → Score 2.0/5
**Buffett says "really cheap," Li Lu says "uncertain, so don't buy"**—such conflict is the real state of investment decisions. Single prompts cannot create such multi-perspective confrontation, which is key to avoiding blind spots.
### 3. Structured Anti-Bias Mechanism
The most dangerous thing about AI is not giving wrong answers but giving **answers that seem correct but are not rigorous enough**. AI Berkshire has built-in multi-layer "fraud prevention" mechanisms:
| Mechanism | Solves What Problem | Example |
|------|------------|------|
| **Information Richness Rating (A/B/C)** | Prevents the illusion that "more data = higher certainty" | Bubble Mart rated B: limited data, estimated indicators marked with confidence level |
| **Munger-style Reverse Testing** | Forces thinking of failure scenarios | "What if Pinduoduo dies?" → List 5 major scenarios and probabilities |
| **Quick Veto List** | 8 red lines for immediate veto | Management integrity issues → Veto immediately, regardless of how cheap valuation is |
| **Anti-Consensus Check** | Avoids aligning with market views | "Why are smart people shorting?" → Discover overlooked risks |
| **Leave Blank Principle** | Rather say "don't know" | Data insufficient, mark "gray area," do not speculate certainty |
### 4. Financial Data Accuracy
LLM mental arithmetic is unreliable. A wrong decimal point in PE or confusing HKD and RMB units can lead to incorrect investment decisions.
**Real Case**: When analyzing Tencent, different sources had market value data in "HKD billion" and "RMB billion" units. AI Berkshire's approach:
```bash
# Market cap manual verification: share price × total shares, compare with report data
python3 tools/financial_rigor.py verify-market-cap \
--price 510 --shares 9.11e9 --reported 4.65e12 --currency HKD
# ✅ Verification passed, deviation only 0.08%
```
All calculations use Python `decimal.Decimal` (precise decimal), not `float`. Key data cross-verified from at least 2 independent sources.
### 5. Reproducible Research Process
Directly asking AI, each output has different formats, depths, and coverage—today analyzing Tencent has a moat score, tomorrow analyzing Meituan might forget.
AI Berkshire ensures: **same input → consistent output in structure and depth**. This means you can:
- Compare 7 companies horizontally with completely consistent scoring standards
- Re-analyze the same company after half a year and directly compare changes
- Align research results among team members
> Real output—7 companies screened with the same standard checklist:
>
> | Company | Pass? | Capability Circle | Good Business | Moat | Management | Margin of Safety | Comprehensive |
> |------|:-----:|:------:|:------:|:------:|:------:|:-------:|:----:|
> | Moutai | ✅ Pass | ★★★★★ | ★★★★★ | ★★★★★ | ★★★☆☆ | ★★★★☆ | 4.7 |
> | Tencent | ✅ Pass | ★★★★☆ | ★★★★★ | ★★★★★ | ★★★★★ | ★★★★☆ | 4.7 |
> | NVIDIA | ✅ Conditional | ★★★★☆ | ★★★★★ | ★★★★★ | ★★★★★ | ★★★☆☆ | 4.3 |
> | Meituan | ✅ Conditional | ★★★★☆ | ★★★★☆ | ★★★★☆ | ★★★★☆ | ★★★★☆ | 4.0 |
> | Kuaishou | ✅ Conditional | ★★★☆☆ | ★★★★☆ | ★★★★☆ | ★★★★☆ | ★★★★★ | 4.0 |
> | Pinduoduo | ❓ Gray | ★★★★☆ | ★★★★☆ | ★★★☆☆ | ★★★☆☆ | ★★★★★ | 3.8 |
> | Bubble Mart | ❓ Gray | ★★★☆☆ | ★★★★☆ | ★★★★☆ | ★★★★★ | ★★★☆☆ | 3.7 |
### 6. Multi-Agent Parallel = Research Depth Doubled
`/investment-team` starts 4 independent Agents **simultaneously** researching one company. Each Agent searches, cross-verifies data, and gives independent conclusions. This is not splitting one prompt into four sections—4 "analysts" each do complete research, and the Team Lead synthesizes.
One person directly asking AI has one context window. 4 Agents in parallel equal 4 times the search volume, 4 times the information sources, and 4 independent perspectives.
```
┌─────────────────────────────────────────────┐
│ Team Lead (You) │
│ Coordinate · Summarize │
├──────┬──────┬──────────┬───────────┤
│ Agent 1 │ Agent 2 │ Agent 3 │ Agent 4 │
│ Business Model │ Financial Valuation │ Industry Competition │ Risk Management Layer │
│ Duan Yongping │ Warren Buffett │ Charlie Munger │ Li Lu Perspective │
└──────┴──────┴──────────┴───────────┘
↓ Parallel Research, Real-time Progress Report ↓
Final Comprehensive Report
```
### Summary
> **Ordinary people asking AI get "看起来对的分析" (看起来对的分析, lit. "analysis that looks correct"), while AI Berkshire provides "可以拿来做决策的投研报告" (投研报告, lit. "research report that can be used for decision-making").**
---
## Overall Architecture
<p align="center">
<img src="assets/architecture.png" alt="AI Berkshire Overall Architecture" width="600" />
</p>
> Source: [`assets/architecture.mmd`](assets/architecture.mmd) (Mermaid editable source code)
**Three-Layer Design Philosophy**:
- **Skill Layer**: Abstract "what you do" into 16 clear entries—depth research, financial analysis, industry screening, position management, thinking tools, select by scenario.
- **Agent Layer**: Each skill internally has 4 Agents in parallel—they independently search, judge, and challenge each other, finally synthesized by Team Lead.
- **Tool Layer**: Precise calculation, real-time retrieval, report inspection—guarantee data rigor and verifiability.
---
## Skills Overview (16)
### 🔬 Depth Research
| Skill | Purpose | Suitable Scenarios |
|-------|------|---------|
| [`/investment-research`](skills/investment-research.md) | Comprehensive analysis by 4 masters | For comprehensive investment research on a listed company |
| [`/investment-team`](skills/investment-team.md) | Multi-Agent parallel research team | 4 Agents researching simultaneously, fastest and most comprehensive |
| [`/management-deep-dive`](skills/management-deep-dive.md) | In-depth management research | When management is a core variable, deep dive |
| [`/private-company-research`](skills/private-company-research.md) | Unlisted company in-depth research | Research on information-scarce unlisted companies like Ant, SpaceX |
| [`/deep-company-series`](skills/deep-company-series.md) | 8-article series dissecting a company | Public account-level in-depth series, 120,000 words from cognitive reset to decision-making loop |
### 📊 Financial Analysis
| Skill | Purpose | Suitable Scenarios |
|-------|------|---------|
| [`/earnings-review`](skills/earnings-review.md) | Financial report reading (primary materials) | Only read original financial reports, not dependent on second-hand research reports, like Buffett reading annual reports |
| [`/earnings-team`](skills/earnings-team.md) | Financial report reading team + public account release | 4 masters interpret financial reports → editing and polishing → reader review → publishable articles |
### 🏭 Industry Screening
| Skill | Purpose | Suitable Scenarios |
|-------|------|---------|
| [`/industry-research`](skills/industry-research.md) | Industry chain panoramic scan | Research all investment opportunities in an industry (slice by industry chain links) |
| [`/industry-funnel`](skills/industry-funnel.md) | Industry funnel screening | Whole market → rough screen ≤10 companies → final selection 3 companies for in-depth analysis |
| [`/quality-screen`](skills/quality-screen.md) | De-bias screening (7 hard indicators) | Quickly exclude non-first-class companies, support stock/industry/index/theme batch screening |
| [`/investment-checklist`](skills/investment-checklist.md) | Buffett's pre-buy checklist | 6 steps quickly screen, 10 minutes decide if worth in-depth |
### 📈 Position Management
| Skill | Purpose | Suitable Scenarios |
|-------|------|---------|
| [`/portfolio-review`](skills/portfolio-review.md) | Portfolio management and optimization | From "research companies" to "manage portfolio"—positioning, concentration, rebalancing |
| [`/thesis-tracker`](skills/thesis-tracker.md) | Investment thesis tracking | Buy-in discipline system: continuously track if thesis is disproven |
| [`/news-pulse`](skills/news-pulse.md) | Stock price abnormality quick attribution | Stock price surged/dropped, 10 minutes understand "what happened" |
### 🧠 Thinking Tools
| Skill | Purpose | Suitable Scenarios |
|-------|------|---------|
| [`/dyp-ask`](skills/dyp-ask.md) | Duan Yongping Q&A | Thinking with Duan Yongping's perspective—business, investment, life |
| [`/financial-data`](skills/financial-data.md) | Financial data acquisition and cross-verification specifications | Ensure key data from 2 independent sources, error >1% alert |
---
## Quick Start
### 1. Install Claude Code
```bash
npm install -g @anthropic-ai/claude-code
```
### 2. Install Skills
Copy `.md` files in `skills/` directory to your Claude Code commands directory:
```bash
# Clone repository
git clone https://github.com/xbtlin/ai-berkshire.git
# Copy skills to Claude Code global commands directory
cp ai-berkshire/skills/*.md ~/.claude/commands/
```
### 3. Usage
Call directly in Claude Code:
```bash
# In-depth research
/investment-research Tencent
/investment-team Meituan
/management-deep-dive Wang Xing Meituan
/private-company-research SpaceX
/deep-company-series Pinduoduo
# Financial analysis
/earnings-review Tencent 2025Q4
/earnings-team PDD 2025 annual report
# Industry screening
/industry-research Nuclear Power
/industry-funnel AI computing power
/quality-screen Hang Seng Index components
/investment-checklist Moutai, NVIDIA, Apple
# Position management
/portfolio-review Tencent 30%, Meituan 20%, Moutai 20%, cash 30%
/thesis-tracker Pinduoduo
/news-pulse Tencent
# Thinking tools
/dyp-ask What is Pinduoduo's moat?
```
---
## Detailed Introduction to Each Skill
### 1. `/investment-research` — Comprehensive Analysis by 4 Masters
Most comprehensive single-company in-depth research framework. Execute in 7 modules:
```
Data collection → Business nature (Duan Yongping) → Moat (Warren Buffett) → Contrarian thinking (Charlie Munger)
→ Management assessment (Duan Yongping + Warren Buffett) → Civilization trend (Li Lu) → Valuation and margin of safety
```
**Core Features**:
- AI research bias self-awareness mechanism (A/B/C level information richness rating)
- Key data multi-source cross-verification (market cap manual verification, at least 2 independent sources)
- 4 masters' "questioning" throughout
- 3 scenarios valuation (optimistic / neutral / pessimistic) + reverse DCF
**Output Example Excerpt**:
> #### Comprehensive Decision Memorandum
>
> | Dimension | Conclusion | Confidence |
> |------|------|--------|
> | Business Quality (Duan Yongping) | Excellent: platform-type business, double-sided network effects, marginal cost tends to zero | ★★★★★ |
> | Moat (Warren Buffett) | Wide and widening: network effects + conversion costs + scale effects triple overlap | ★★★★☆ |
> | Management (Duan Yongping + Warren Buffett) | Excellent: founder-led, strong capital allocation discipline | ★★★★☆ |
> | Max Risk (Charlie Munger) | Regulatory policy uncertainty, new business loss drags overall profit | ★★★☆☆ |
> | Civilization Trend (Li Lu) | Aligns with digital consumption trend, but not "civilization-level paradigm shift" | ★★★★☆ |
> | Valuation (Warren Buffett + Duan Yongping) | Current PE 18x, historical median low, certain margin of safety | ★★★★☆ |
>
> **Duan Yongping**: "The nature of this business is connecting consumers and merchants, making money from efficiency improvements. Good business signs: users increase, merchants increase; merchants increase, users increase. The flywheel once starts, hard to stop."
> **Munger**: "Turn it around—if this company disappears tomorrow, users and merchants will do what? If the answer is 'they will find alternatives quickly', then the moat is not deep enough. If the answer is 'life will become very inconvenient', then it is worth attention."
---
### 3. `/investment-checklist` — Buffett's Pre-Investment Checklist
A quick six-step screening process to help you decide if a company is worth in-depth research within 10 minutes:
```
Step 1: Circle of Competence (Do I understand it?)
↓ Pass
Step 2: Good Business (What are its economic characteristics?)
↓ Pass
Step 3: Moat (How strong is its competitive advantage?)
↓ Pass
Step 4: Management (Is it trustworthy?)
↓ Pass
Step 5: Margin of Safety (Is the price cheap?)
↓ Pass
Step 6: Decision Discipline (Is it rational or FOMO?)
↓ Pass
✅ Mirror Test
```
**Supports multi-company comparison** — Screen multiple stocks at once:
```
/investment-checklist Tencent, Alibaba, Meituan, Pinduoduo
```
**Sample output excerpt**:
> #### Mirror Test
>
> "I bought Tencent at 380 HKD because:
> 1. The business nature is **social network + digital content platform**, which I understand;
> 2. Its moat is **the social relationship chain of 1.2 billion users**, and it's getting wider;
> 3. The management team **Pony Ma is low-key and pragmatic, with excellent capital allocation**, which is trustworthy;
> 4. The current price is equivalent to **80% of its intrinsic value**, which has a certain margin of safety;
> 5. Even if I'm wrong, the downside risk is controllable because **the net cash on its books exceeds 200 billion, and the game cash flow is strong**."
>
> ✅ Pass the mirror test
>
> **If you can't complete 5 sentences, don't buy. No exceptions.**
---
### 4. `/industry-research` — Industry Chain Research
Starting from an investment theme, complete an industry chain research:
```
Build investment logic chain → Draw industry chain panorama → Scan global listed companies
→ Analyze top companies in each link → Investment portfolio configuration suggestions
```
**Sample output excerpt**:
> #### Investment Logic Chain: Nuclear Power
>
> Underlying trend: AI data center power demand surge + Carbon neutrality target
> → Leads to: Stable and clean base load power demand surge
> → Creates: Nuclear power restart/new construction/SMR's deterministic demand
> → Benefits: Uranium mining → Fuel processing → Equipment manufacturing → Operators
>
> #### Recommended Portfolio
>
> | Level | Position | Target | Link | Core Logic |
> |------|------|------|------|---------|
> | Core | 50% | CGN, Cameco | Operation + Uranium mining | Highest certainty |
> | Satellite | 30% | China Nuclear Power, Dongfang Electric | Operation + Equipment | Benefiting from domestic substitution |
> | Option | 15% | NuScale, Nano Nuclear | SMR | High risk, high elasticity |
> | ETF | Alternative | URA, URNM | Full chain | Lazy solution |
---
### 5. `/industry-funnel` — Industry Funnel Screening
Starting from an industry/direction, **from the whole market → ≤10 companies → 3 companies**, layer by layer screening:
```
Full market scan (activity + increase + top 30 market value, 30-60 companies)
↓ 5 strict value investment indicators
Coarse screening ≤ 10 companies
↓ Fine analysis (300-500 words per company)
Fine analysis ≤ 10 companies
↓ Final selection (based on portfolio complementarity, not top 3 scoring)
In-depth analysis of top 4 masters for 3 companies (800-1200 words per company)
↓
Recommended portfolio (core/satellite/option) + Operation signals
```
**Core features**:
- Each layer has clear standards for retention or elimination, and eliminated targets leave reasons for elimination (not a black box)
- The final 3 companies are selected based on "portfolio complementarity" (high certainty + medium elasticity + high elasticity), not the top 3 scoring
- Forced listing of "future IPO candidates" to avoid missing core players in the primary market
- AI bias self-examination mechanism: to respond to leader preference/English preference/story preference/listings preference
**Difference from `/industry-research`**:
- `industry-research` focuses on industry chain structure and panorama (sliced by link)
- `industry-funnel` focuses on individual stock screening funnel (layer by layer screening from the whole market to 3 companies)
**Actual test: AI industry 4 sub-track parallel (2026-05-09)**:
| Sub-track | Final 3 companies | Core position recommendation |
|-------|---------|------------|
| AI computing power | TSMC / NVIDIA / SK Hynix | TSMC ★★★★★ |
| AI model | Alphabet / Meta / Alibaba | Alphabet ★★★★★ |
| AI application | Microsoft / Adobe / AppLovin | Microsoft + Adobe ★★★★ |
| AI infrastructure power | Eaton / TBEA / Talen Energy | Eaton + TBEA ★★★★ |
**Key findings**: The biggest winner in the AI application layer is not an AI-native company, but a mature giant with channel + data + workflow embedding degree - this echoes the historical pattern of "selling shovels" in the 1995-2000 internet bubble (Amazon and Apple won, Pets.com lost).
Complete report: [AI computing power](reports/AI%20computing%20power-funnel-20260509.md) · [AI model](reports/AI%20model-funnel-20260509.md) · [AI application](reports/AI%20application-funnel-20260509.md) · [AI infrastructure power](reports/AI%20infrastructure%20power-funnel-20260509.md)
---
### 6. `/private-company-research` — Private Company Research
A "detective-style" research framework designed for private companies with scarce information:
**Core differentiation**:
- **Financial data piecing**: Puzzle financial data from prospectus, parent company financial report, financing news, and industry data
- **Confidence level marking**: Mark each data point with 🟢high / 🟡medium / 🔴low confidence level
- **Multi-method valuation crossover**: Financing valuation method + comparable company method + DCF + terminal reverse deduction method
- **Exit path analysis**: IPO/merger/secondary transfer path assessment
**Sample output excerpt**:
> #### Company profile: SpaceX
>
> | Project | Content |
> |------|------|
> | Latest valuation | ~$350B (2025 secondary market) 🟡 |
> | Estimated revenue | ~$13 billion (2024) 🟡 |
> | Starlink users | 4 million+ (end of 2024) 🟢 |
> | Launch frequency | 100+ times/year (2024) 🟢 |
>
> #### Valuation judgment
>
> | Method | Valuation range | Explanation |
> |------|---------|------|
> | Recent financing | $350B | Secondary market quotation, with liquidity premium |
> | Comparable company method | $200-280B | Benchmarking telecom + aerospace + defense |
> | DCF (neutral) | $250-350B | Assuming Starlink $3 billion revenue in 2027 |
> | Terminal reverse deduction | $400-600B | Assuming Starlink becomes global telecom infrastructure |
>
> **Comprehensive reasonable valuation range: $250B - $400B**
---
### 7. `/news-pulse` — Stock Price Movement News Attribution
A situation response skill designed for quickly understanding what happened during stock price movements **(not in-depth research, 10-15 minutes quick attribution)** — to avoid anxiety or blind stop loss when holding positions change.
**Core differentiation**:
- **4-dimensional parallel reconnaissance**: Company events / regulatory policies / industry opponents / market sentiment (sell-side + big V + southbound funds)
- **Attribution prioritizes over listing**: Not listing all news, but judging "which event deserves this stock price movement"
- **Forced nature judgment**: Value event / emotional fluctuation / **unknown reason** / mixed — where "unknown reason" is the most valuable output (may have insider scoop)
- **Clear action suggestion**: Whether to trigger in-depth research, whether to review thesis, or just observe
**Difference from other skills**:
| Scenario | What to use |
|------|------|
| Complete investment research (hour level) | `/investment-team` or `/investment-research` |
| In-depth financial report reading | `/earnings-review` |
| Long-term thesis tracking | `/thesis-tracker` |
| **Stock price movement 10-minute attribution** | **`/news-pulse`** |
**Sample output excerpt** (Tencent 4/17-5/01 actual test, 14 days -10.47%):
> #### One-sentence attribution
> This -10.47% drop is about 70-80% driven by capital and sentiment (repurchase silent period + southbound capital reduction + sector beta + AI narrative being taken away), and 20-30% by deferred digestion of doubled AI investment — **no fundamental negative**, sell-side maintains buy consensus, which is a "liquidity + sentiment-driven" callback, not a value event.
>
> #### Abnormal attribution table
>
> | Candidate explanation | Estimated contribution | Confidence level |
> |---------|--------|--------|
> | Disappearance of repurchase silent period (structural, before 5/13 financial report) | -3% ~ -4% | High |
> | Southbound funds turning to net sell Tencent | -2% ~ -3% | High |
> | AI narrative being taken away by competitors (DeepSeek V4/Qwen3.6/month dark 1T) | -1% ~ -2% | Medium |
> | Sector/macro beta (oil price + geopolitics + Fed Warsh hawkish) | -2% ~ -3% | High |
> | Pre-earnings caution | -1% ~ -2% | Medium |
> | Fundamental deterioration | **0%** | Extremely high (excluded) |
>
> #### Nature judgment: ✅ Mixed
> 70% capital/sentiment + 20% AI long-term narrative concern + 10% pre-earnings uncertainty
>
> **Key rebuttal**: D段永平 4/8 selling Tencent put (bullish); sell-side 24-house consensus Strong Buy; NetEase 4/30 up 2% against the market (excluding game industry issues); Tencent ran 7 percentage points worse than Hengke (Hengke actually rose 4% this month).
Invocation method:
```
/news-pulse Tencent
/news-pulse Pinduoduo down 12% within one week
/news-pulse miHoYo
```
---
## Practical Research Reports
> The following are real investment research reports generated using this framework, showcasing the actual output effect of AI investment research.
| Company | Used skill | Core conclusion | Report link |
|------|-----------|---------|---------|
| Pinduoduo (PDD) | `/investment-team` | Comprehensive 3.4/5, extremely cheap but insufficient 10-year certainty, suitable for medium position | [View report](reports/Pinduoduo/) |
| Tencent Holdings (0700.HK) | `/investment-research` | Social monopoly + excellent capital allocation, 14x forward PE reasonable and low | [View report](reports/Tencent/) |
| 7 company comparison | `/investment-checklist` | Moutai and Tencent passed; NVIDIA, Meituan, and Kuaishou conditionally passed; Pinduoduo and Pop Mart gray | [View report](reports/Multi-company%20comparison-checklist-20260408.md) |
| Master holding tracking | Custom research | Latest 13F holdings of Buffett/Lu Zhong/Li Xueqin + Pinduoduo cost analysis | [View report](reports/Master%20holding%20tracking-research-20260408.md) |
> *More reports will be added continuously. Welcome to submit your research reports generated using this framework.*
---
## Design Concept
### Four Masters' Methodology Fusion
```
┌──────────────────┐
│ Duan Yongping │
│ "Right business" │
│ Business model essence│
└────────┬─────────┘
│
┌──────────────────┼──────────────────┐
│ │ │
▼ ▼ ▼
┌────────┐ ┌──────────┐ ┌────────┐
│ Buffett │ │ Munger │ │ Li Xueqin │
│ Moat │ │ Reverse thinking │ │ Civilization trend│
│ Margin of safety│ │ Risk list │ │ Paradigm shift│
│ Management │ │ Bias self-examination │ │ Industry value│
└────────┘ └──────────┘ └────────┘
```
The four masters are not simply divided, but designed to **challenge each other**:
- Duan Yongping says "good business", Munger will ask "how will it die"
- Buffett says "cheap enough", Li Xueqin will ask "will it still be there in 10 years"
- What you get is not a simple combination of four reports, but a collision of four thinking modes
### Financial Rigor Tool (`tools/financial_rigor.py`)
| Function | Command | Problem solved |
|------|------|-----------|
| **Market capitalization calculation** | `verify-market-cap` | Accurate calculation of stock price × total shares, detecting unit errors |
| **Valuation calculation** | `verify-valuation` | Accurate decimal calculation of PE/PB/ROE/FCF Yield |
| **Multi-source cross-validation** | `cross-validate` | Automatic comparison of the same data from N sources, alerting if exceeding tolerance |
| **Three-scenario valuation** | `three-scenario` | Accurate calculation of target price under optimistic/neutral/pessimistic scenarios |
| **Benford's Law detection** | `benford` | Detection of abnormal first-digit distribution in financial data |
| **Accurate calculator** | `calc` | Accurate calculation of arbitrary financial expressions, replacing LLM mental calculation |
**Design principle**: All calculations use Python `decimal.Decimal` (accurate decimal), not `float` (floating-point approximation). `0.1 + 0.2 = 0.3` is not allowed to fail in financial scenarios.
---
## Project Roadmap
- [x] Four masters' comprehensive analysis framework (`/investment-research`)
- [x] Multi-agent parallel investment research team (`/investment-team`)
- [x] Buffett's pre-investment checklist (`/investment-checklist`)
- [x] Industry chain panorama scanning (`/industry-research` + `/industry-funnel`)
- [x] Private company research framework (`/private-company-research`)
- [x] Financial rigor tool (accurate arithmetic, market capitalization verification, multi-source cross-validation, Benford's Law detection)
- [x] Stock price movement rapid attribution (`/news-pulse` 4-dimensional parallel reconnaissance)
- [x] Financial report in-depth reading (`/earnings-review` + `/earnings-team` four masters parallel interpretation)
- [x] Investment portfolio management (`/portfolio-review` position review and rebalancing)
- [x] Investment thesis tracking (`/thesis-tracker` post-buying discipline system)
- [x] Management in-depth research (`/management-deep-dive`)
- [x] Quick elimination screening (`/quality-screen` 7 strict indicators)
- [x] Duan Yongping thinking simulation (`/dyp-ask`)
- [x] In-depth series long article (`/deep-company-series` 8 articles, 120,000 words)
- [ ] Historical backtest: AI research report vs. actual stock price performance
- [ ] Macroeconomic cycle analysis framework
- [ ] Real-time data access based on MCP (Wind/Bloomberg/Yahoo Finance)
---
## Disclaimer
This project is for learning and research purposes only and does not constitute any investment advice. Investment carries risks, and decisions need to be made with caution. Please always conduct your own due diligence (DYOR).
---
## Star History
If this project helps you, give it a Star!
## License
MIT License
---
> "The best investment you can make is in yourself." — Warren Buffett
>
> AI Berkshire: Let everyone have their own investment research team.
