Codex adapter note
This skill is generated from skills/management-deep-dive.md so Claude Code and Codex users share one canonical workflow.
- Treat
$ARGUMENTSas the user's request in the current Codex thread. - When the source mentions Claude-only surfaces such as Task, Agent, WebSearch, Bash, Read, or Write, use the closest Codex capability available in this session: subagents when available, web search when needed, shell commands for local tools, and normal file edits for workspace files.
- Use shared project tools from
tools/in this repository. Prefer running commands from the repository root with paths likepython3 tools/financial_rigor.py ...; if the current thread starts outside the repo, locate the actual checkout path first instead of assuming a fixed home-directory path. - Before starting research, run the
datecommand to confirm today's date; treat it as the baseline for "latest" data and state the data cutoff date in the report header. Never assume the current date from training data. - Preserve the research quality rules from
AGENTS.md: cross-check financial data, use exact arithmetic tools for valuation/math, and clearly label uncertainty and source gaps.
Conduct in-depth management research on $ARGUMENTS.
Supported input formats：公司名 or 人名 公司名, for example:美团、王兴 美团、黄仁勋 英伟达
"Buying stocks is buying people. Find someone you trust, then hold for the long term." — Duan Yongping
"To evaluate management, look at what they do when no one is watching." — Buffett
Design Philosophy
Most investment analysis evaluates management only superficially: résumé, shareholding percentage, compensation. But Buffett spends a lot of timeeating and chatting with management, Li Lu saysthe essence of his investing is investing in people, and Duan Yongping saysbuying stocks is buying people。
This Skill is the /investment-research fifth step, management evaluation, of thedeepened version. When the management score in standard investment research is uncertain (★★★ or below), or management is the core investment thesis, use this Skill for in-depth research.
AI cannot eat meals with management, but through public information it can:
- Track whether management's words and actions are consistent(promises vs. delivery)
- Analyze the returns of every major capital allocation decision
- Infer character from decisions made during difficult times
- Cross-verify through feedback from employees/merchants/customers
Execution Process
Step One: Identify key management and launch parallel data collection
Use WebSearch to confirm the following key people:
| Role | Name | Tenure | Background | Shareholding/Options |
|---|---|---|---|---|
| CEO/Chairman | ||||
| CFO | ||||
| Founder (if not in office) | ||||
| Actual controller (if different from CEO) | ||||
| Other key executives |
Note: Distinguish "who makes the decisions" from "whose name is on the title." Some company founders, though retired, are still the soul of the company (such as Huang Zheng at Pinduoduo).
After confirming the key people, use the Task tool to launch multiple background Agents in parallelto collect the following data:
- Agent 1: CEO public statements and forecast record (shareholder letters, earnings calls, interviews, social media)
- Agent 2: Capital allocation decision record (M&A, buybacks, dividends, new business investments)
- Agent 3: Governance structure and compensation (equity structure, related-party transactions, executive compensation)
- Agent 4: Cross-verification information (employee reviews, customer feedback, industry reputation)
Step Two: CEO Circle of Competence Assessment
2.1 Strategic Vision
Search the CEO's public statements over the past 5 years (shareholder letters, earnings calls, interviews, social media), and extract their judgments on the following questions:
| Time | CEO's judgment/forecast | Actual result | Accuracy |
|---|---|---|---|
| "We believe the X market will..." | The X market actually... | ✅/❌ | |
| "Our focus over the next 3 years is..." | Actual execution... | ✅/❌ |
Key question：
- Has the CEO ever made a correct judgment ahead of the market?
- Has the CEO remained calm when everyone was optimistic?
- Is the CEO's understanding of industry trends following the market or independent thinking?
2.2 Execution Capability
| Dimension | Assessment | Evidence |
|---|---|---|
| From strategy to implementation | Did they do what they said they would do? | |
| Organizational capability | Can they attract and retain talent? | |
| Crisis management | How do they respond when facing difficulties? | |
| Iteration speed | Is the speed of correction after mistakes fast? |
Step 3: Integrity Assessment (Most Important)
Buffett: "We look for three qualities: integrity, intelligence, and energy. If you don't have the first, the other two will kill you."
3.1 Commitment vs. Delivery Tracking
From the past 3 years of earnings calls, shareholder letters, and public interviews, extract thespecific commitments：
| # | Time | Commitment content | Commitment occasion | Delivery status | Assessment |
|---|---|---|---|---|---|
| 1 | "We will achieve profitability in business X by 2025" | 2024 annual report earnings call | ✅/⚠️/❌ | ||
| 2 | "We plan to repurchase $X billion" | 2024 shareholder letter | ✅/⚠️/❌ |
Delivery rate statistics：
| Commitment delivery rate | Assessment |
|---|---|
| >80% | Excellent — does what they say |
| 60-80% | Acceptable — generally right direction but execution has deviations |
| 40-60% | Concerning — overpromises and underdelivers |
| <40% | Serious problem — untrustworthy |
3.2 Performance During Difficult Times
Search for major crises/difficulties the company has experienced in its history (stock price plunge, performance miss, regulatory shock, intensified competition), and analyze management's response:
| Crisis event | Time | Management response | Assessment in hindsight |
|---|
Focus：
- Did they communicate proactively or avoid?
- Did they attribute it internally or shift blame externally?
- Did they take the opportunity to do the difficult but right thing, or choose to please the market in the short term?
3.3 Attitude Toward Stakeholders
| Stakeholder | Management attitude | Evidence | Assessment |
|---|---|---|---|
| Shareholders | Respect/ignore/exploit | ||
| Employees | Treat well/exploit/ignore | ||
| Customers/users | Customer-centric/short-term extraction | ||
| Merchants/suppliers | Fair cooperation/extreme price suppression | ||
| Regulators/society | Compliance and cooperation/playing edge ball |
Li Lu: "Attitude toward stakeholders determines a company's long-term vitality. Short-term extraction can improve efficiency, but in the long term it damages the ecosystem."
Step 4: Capital Allocation Capability
This is the management capability Buffett values most —for every dollar earned, how many dollars can management turn it into?
4.1 Capital Allocation Decision Record
Search for the company's major capital allocation decisions over the past 5 years, and evaluate each one:
M&A record：
| Time | Acquisition target | Amount | Strategic Logic | Ex-post Return | Score (1-5) |
|---|
Buyback Records：
Use tools/financial_rigor.py verify-valuation to verify valuation metrics such as PE at the time of buyback and currently.
| Time | Buyback Amount | Average Buyback Price | PE at the Time | Ex-post Review | Score (1-5) |
|---|
Dividend Records：
| Year | Dividend Amount | Dividend Payout Ratio | FCF for the Same Period | Sustainable |
|---|
New Business Investment：
| Time | Investment Field | Cumulative Investment | Current Status | Return Assessment | Score (1-5) |
|---|
4.2 Capital Allocation Score
| Dimension | Score (1-5) | Notes |
|---|---|---|
| M&A Discipline | Acquisitions at reasonable prices? How is post-acquisition integration? | |
| Buyback Timing | Buybacks when undervalued, stopped when overvalued? | |
| Dividend Reasonableness | Does the dividend payout ratio match FCF? | |
| New Business Investment | What is the success rate? How is stop-loss discipline? | |
| Cash Management | Is the cash reserve reasonable? Is too much being hoarded? | |
| Overall Score |
Buffett Standard: An ideal management invests decisively when there are good opportunities, actively buys back shares/pays dividends when there are no good opportunities, and never makes overpriced acquisitions.
Step Five: Governance Structure Assessment
5.1 Equity Structure
| Item | Details | Risk Assessment |
|---|---|---|
| Are there AB shares/super-voting rights? | ||
| Founder/actual controller shareholding percentage? | ||
| Is there a VIE structure? | ||
| Are independent directors truly independent? | ||
| Recent increase/decrease in holdings by major shareholders? |
5.2 Compensation Reasonableness
| Executive | Annual Total Compensation | As % of Company Net Profit | Compared with Peers | Reasonable |
|---|
Focus: Is the incentive structure aligned with long-term shareholder interests? Or does it encourage short-term behavior?
5.3 Related-Party Transactions
| Related Party | Transaction Content | Amount | Fair or Not | Risk Assessment |
|---|
Step Six: Side Verification
AI cannot communicate face-to-face with management, but can verify through side information from public channels.Note: The following information depends on publicly searchable content and may be incomplete; indicate the information source and availability.
6.1 Employee Perspective
Search Glassdoor rating summaries, Zhihu discussions, etc.Publicly searchableemployee reviews (for platforms requiring login such as Maimai, mark "users may supplement on their own"):
| Dimension | Rating Trend | Key Feedback |
|---|---|---|
| Corporate Culture | ||
| Management Evaluation | ||
| Work Intensity | ||
| Compensation Satisfaction | ||
| Growth Prospects |
6.2 Customer/Merchant Perspective
Search App Store ratings, consumer complaints, merchant forums:
| Dimension | Rating/Trend | Key Feedback |
|---|---|---|
| Product Satisfaction | ||
| Customer Service | ||
| Merchant/Supplier Relationships |
6.3 Industry Reputation
Search industry forums and social media to understand peers' and industry insiders' evaluations of this management team.
Step Seven: Scenario Analysis After the CEO Leaves
Buffett: "A good company should be one that any fool can run—because sooner or later, a fool will."
| Question | Answer |
|---|---|
| If the CEO leaves tomorrow, can the company operate normally? | |
| How deep is the existing management team? Is there a clear successor? | |
| Does the company's competitive advantage depend on the CEO personally, or on the organization/system? | |
| Have historical management transitions been smooth? |
Step Eight: Output the Management Evaluation Report
Report Structure
一、关键人物速览（表格）
二、诚信度评估
- 承诺兑现率
- 困难时期表现
- 对利益相关方态度
三、能力评估
- 战略眼光（预判准确度）
- 执行能力
- 资本配置记录
四、治理结构
- 股权结构风险
- 薪酬合理性
- 关联交易
五、侧面验证
- 员工视角
- 客户/商家视角
六、综合评分与结论
Overall Score
| Dimension | Weight | Score (1-5) | Weighted |
|---|---|---|---|
| Integrity | 35% | ||
| Strategic and Execution Capability | 25% | ||
| Capital Allocation Capability | 25% | ||
| Governance Structure | 15% | ||
| Overall Score | 100% |
Duan Yongping's "Buying People" Criteria
Answer the following three questions:
- Is this person upright?(Honest, does not take advantage of shareholders)
- Is this person capable?(Strategic vision + execution + capital allocation)
- Are you willing to give your money to this person to manage for 10 years?
All three are "yes" = ★★★★★ (5 points) The first two are "yes" = ★★★★ (4 points) Only the first is "yes" = ★★★ (3 points) The first is not "yes" = ★ (1 point, do not invest)
Step Nine: Save the Report
Write the report to reports/{公司名}-management-{YYYYMMDD}.md, for example reports/美团-management-20260409.md
Key Principles
- Integrity is a veto item — Insufficient ability can be learned; character problems cannot be fixed
- Look at actions, not words — What management says doesn't matter; what they do matters
- See the truth in hardship — When the wind is at their back, everyone is a good CEO; only when the wind is against them do you see true skill
- Capital allocation is the ultimate exam — Making money is easy; allocating the money you make well is hard
- Don't fall in love with management — Stay objective; even someone you admire can make big mistakes