How to Use the Buffett-Style Investment Checklist in AI Berkshire: Complete Implementation Guide

The Buffett-style investment checklist in AI Berkshire is implemented as a declarative Skill at /investment-checklist that executes a six-gate pipeline through skills/investment-checklist.md, producing rigor-checked investment reports using precision tools from tools/financial_rigor.py.

AI Berkshire is an open-source framework that codifies Warren Buffett’s investment philosophy into reproducible AI workflows. The repository’s investment-checklist skill transforms qualitative value investing principles into a deterministic, data-verified analysis pipeline that eliminates float-drift errors and ensures reproducible results.

What Is the Buffett-Style Investment Checklist?

The AI Berkshire investment checklist is a Skill defined in skills/investment-checklist.md that mirrors Buffett’s "buy-before-checklist" methodology. Unlike generic stock screeners, this implementation enforces financial rigor through a six-gate decision pipeline: ability-circle → good-business → moat → management → margin of safety → discipline.

When invoked, the skill spawns parallel Task agents for data gathering, then processes each ticker through the six gates using Python’s decimal.Decimal for exact arithmetic, eliminating floating-point errors that plague traditional financial calculators.

How the Six-Gate Pipeline Works

The checklist evaluates every security through six deterministic gates defined in the skill template:

1. Ability Circle (能力圈)

Determines whether the business falls within the investor’s circle of competence.

2. Good Business (好生意)

Analyzes the underlying economics of the business model and industry dynamics.

3. Economic Moat (护城河)

Evaluates competitive advantages and barriers to entry.

4. Management Quality (管理层)

Assesses capital allocation discipline and shareholder alignment.

5. Margin of Safety (安全边际)

Calculates valuation buffers using three-scenario modeling.

6. Investment Discipline (纪律)

Final gate confirming checklist adherence before position entry.

Running the Investment Checklist Skill

Basic CLI Invocation

Activate the skill through Claude Code or Codex with the slash command:

/investment-checklist 茅台, 英伟达, 苹果

For Codex prompts, use the Chinese invocation format:

使用 investment-checklist 研究 茅台, 英伟达, 苹果

The skill returns a structured table with ★-ratings for each gate:

公司	Checklist通过？	能力圈	好生意	护城河	管理层	安全边际	核心结论
茅台	✅ 通过	★★★★★	★★★★★	★★★★★	★★★☆☆	★★★★☆	高
英伟达	✅ 有条件	★★★★☆	★★★★★	★★★★★	★★★★★	★★★☆☆	中
苹果	✅ 通过	★★★★★	★★★★★	★★★★★	★★★★★	★★★★★	极高

Python SDK Usage

Call the skill programmatically using the Codex CLI:

from pathlib import Path
from subprocess import run, PIPE

def run_checklist(tickers: str):
    cmd = ["codex", "run", "/investment-checklist", tickers]
    result = run(cmd, stdout=PIPE, text=True)
    print(result.stdout)

run_checklist("茅台, 英伟达, 苹果")

Financial Rigor Tools and Validation

The checklist delegates precision calculations to tools/financial_rigor.py, which provides CLI utilities for data verification. These tools ensure the six-gate analysis rests on mathematically exact foundations.

Market Cap Verification

Validate market capitalization against reported figures:

python3 tools/financial_rigor.py verify-market-cap \
  --price 510 --shares 9.11e9 --reported 4.65e12 --currency HKD

Valuation Ratio Validation

Confirm PE, PB, and dividend metrics:

python3 tools/financial_rigor.py verify-valuation \
  --price 510 --eps 23.5 --bvps 120 --fcf-per-share 18 --dividend 2.4

Three-Scenario Valuation

Run Buffett-style conservative, base, and optimistic scenarios:

python3 tools/financial_rigor.py three-scenario \
  --price 510 --eps 23.5 --shares 9.11 \
  --growth 0.15 0.08 0.00 \
  --pe 25 20 15 \
  --currency HKD

These commands produce the exact numbers embedded in the checklist’s final tables, ensuring the margin of safety gate rests on verified data.

Output Format and Reports

Upon completion, the skill generates three artifacts:

• A per-company checklist with ★-ratings for each of the six gates
• A mirror-test statement following the "5-sentence" rule for thesis clarity
• A Markdown report saved to ~/巴菲特Checklist-<name>.md containing the summary table and detailed analysis

Because the checklist is fully declarative, running the same input twice yields identical structured output, enabling reproducible portfolio reviews and thesis tracking.

Summary

• The Buffett-style investment checklist is implemented as the /investment-checklist skill in skills/investment-checklist.md
• Analysis flows through six deterministic gates: ability-circle, good-business, moat, management, margin of safety, and discipline
• Financial rigor tools in tools/financial_rigor.py use decimal.Decimal to eliminate floating-point errors in valuation calculations
• Invoke via Claude Code slash commands, Codex prompts, or Python subprocess calls for batch processing
• Output includes star-rated checklists, mirror-test statements, and persistent Markdown reports

Frequently Asked Questions

How do I customize the six-gate criteria for specific industries?

Edit the prompt template in skills/investment-checklist.md to adjust gate weights or add industry-specific questions. The skill is declarative, so modifications apply consistently across all future runs. Refer to AGENTS.md for the canonical-vs-generated skill architecture if you need to extend the pipeline with custom Task agents.

Can I run the checklist on non-Chinese equities?

Yes. The financial rigor tools support multiple currencies via the --currency flag (e.g., USD, HKD, EUR). Simply pass the ticker symbols in the standard format recognized by your data provider, and the verify-market-cap and three-scenario utilities will handle precision arithmetic regardless of currency denomination.

What makes the financial calculations different from Excel or standard Python floats?

The tools/financial_rigor.py module uses Python’s decimal.Decimal class for all arithmetic operations, which maintains exact precision for financial calculations. This eliminates the float-drift errors common in IEEE 754 floating-point arithmetic, ensuring that valuation ratios and scenario models match audited financial statements to the cent.

Where is the checklist output saved and how can I integrate it with other skills?

Results are automatically saved to ~/巴菲特Checklist-<name>.md in your home directory. The Markdown format is compatible with downstream AI Berkshire skills listed in the README under "Skills 一览". You can pipe the output file path to portfolio-management skills or thesis-trackers for automated workflow chaining.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}