How to Audit a Gates-Inspired Stock Analysis Workflow

A practical way to test business durability, cash flow, valuation, portfolio fit, and the limits of public filings before relying on an AI stock analysis.

By Clover Li | Published 17 July 2026 | Reviewed 23 July 2026

Back to guides About the editor

This $0.99 Codex skill turns a vague stock opinion into a sourced review of business quality, valuation, downside, and portfolio fit.

What the $0.99 product includes

Gates-Style Stock Analysis is a Codex skill for people who want a repeatable way to examine US stocks and portfolios. It asks for current filings, earnings, material news, valuation data, catalysts, and portfolio context before it reaches a conclusion. The output starts with one action: buy gradually, hold, watch, trim, or exit. It then shows the evidence, a score out of 100, bear and bull cases, position-size limits, thesis breakers, and a review date.

Buy the skill for $0.99

The purchase includes the main SKILL.md, an interface file for Codex, two reference documents, and a Python scoring script. Purchase the $0.99 download securely on Gumroad. The package contains no brokerage connection, tracking code, API credential, or personal portfolio data. Buyers receive the ZIP through Gumroad after checkout rather than through a public file link.

Why the name needs a qualification

Bill Gates has not published a complete stock-trading formula. His personal wealth is not the same thing as the Gates Foundation Trust, and the Trust's public filings do not show every asset, hedge, cash balance, purchase price, or decision made after the reporting date. The skill therefore uses the phrase Gates-inspired. It takes public evidence about patient capital allocation and durable businesses as a starting point, then applies ordinary financial analysis. It does not claim that Gates selected a stock or recommends it.

What public records can tell us

The Gates Foundation says its Trust holds and manages investment assets and transfers funds needed for charitable work. The Foundation and Trust are separate entities, and Bill Gates is the trustee. SEC Form 13F filings reveal certain long US positions after each quarter, but they are snapshots rather than live portfolios. Gates has also written about Warren Buffett looking for companies with strong long-term prospects, making substantial investments, and holding through economic cycles. That passage describes Buffett. The skill keeps that attribution intact instead of turning it into a made-up Gates trading rule.

The 100-point score

Business durability, 20 points: demand, disruption risk, and the chance that the company still matters in ten years.
Moat and pricing power, 15 points: switching costs, cost advantage, networks, scarce assets, or another defensible edge.
Free-cash-flow quality, 15 points: conversion, repeatability, capital intensity, and accounting quality.
Balance sheet, 10 points: debt, interest coverage, liquidity, and refinancing pressure.
Management allocation, 10 points: reinvestment, acquisitions, dividends, buybacks, dilution, and candor.
Valuation, 15 points: expected return under conservative assumptions and support for the downside.
Long-term runway, 10 points: credible opportunities to reinvest without heroic forecasts.
Portfolio fit, 5 points: concentration, correlation, and overlap with positions already owned.

How the score is interpreted

A score of 80 or more identifies a possible ownership candidate, but only if the valuation work also passes. A score from 70 to 79 normally belongs on a watchlist or in a small starter position. A score from 60 to 69 needs a specific thesis and modest size. A lower score points toward avoidance, trimming, or exit. These bands are decision aids, not backtested promises. A company can score well and still be too expensive. A cheap stock can also remain cheap because its business is deteriorating.

The valuation gate

The skill requires bear, base, and bull cases. Mature companies are usually checked with free-cash-flow yield, owner earnings, and a simple discounted cash-flow model. Cyclical businesses need normalized earnings rather than one unusually strong year. Young companies need dilution-aware scenarios and evidence that unit economics improve with scale. The base-case return must compete with a realistic lower-risk alternative such as Treasury bills or a broad index fund. If it does not, the correct result may be watch even when the company is excellent.

Portfolio rules are part of the answer

A stock is not attractive in isolation if it makes an already concentrated portfolio harder to survive. The skill normally limits a durable profitable company to roughly 10 to 15 percent of a portfolio. An unprofitable or highly volatile company is usually limited to 2 to 5 percent. Long-option premium is treated as fully at risk and normally stays near 0.5 to 1 percent. Leveraged ETFs require a tactical horizon and an exit rule because daily reset and volatility can damage long holding-period results.

Install on macOS or Linux

Download the ZIP from the purchase receipt and unzip it.
Move the gates-style-stock-analysis folder into ~/.codex/skills/.
Restart Codex so the new skill appears in the skill list.
Invoke it with: Use $gates-style-stock-analysis to analyze MSFT using current data.
For a portfolio review, include ticker, quantity, cost, account size, time horizon, and maximum loss you can tolerate.

Run the score manually

The included script accepts eight scores from zero to ten. For example: python3 scripts/score_stock.py --durability 9 --moat 8 --cash-flow 9 --balance-sheet 8 --management 7 --valuation 6 --runway 8 --portfolio-fit 7. The script calculates the weighted result and prints each contribution. It does not fetch market data. That separation is deliberate: current facts must be checked before a human or agent assigns the inputs.

What a useful request looks like

A weak request says, tell me a stock that will rise sharply. A useful request says, compare Microsoft, Oracle, and Salesforce for a three-year holding period; use current filings and prices; assume a $20,000 portfolio with 20 percent already in Microsoft; show the bear case and do not recommend a position larger than 10 percent. The second request gives the skill enough context to discuss opportunity cost and concentration rather than producing a generic stock pitch.

Limits and financial risk

This skill is an analysis template, not a broker, adviser, signal service, or guarantee. Public data can be wrong, delayed, revised, or misunderstood. Forecasts can fail. Price gaps can exceed an intended exit level, and options can lose the entire premium. Check the cited filing and calculation before trading. Use position sizes that let the portfolio survive an error. If a decision affects money needed for living expenses, taxes, debt repayment, education, or medical care, seek advice from a qualified professional who can review the full situation.

Sources checked

Continue with a related guide

Editorial note: Clover Li reviewed the workflow, examples, links, and changing claims on 23 July 2026. AI may have assisted with outlining or language cleanup, but the editor remains responsible for the published page. Read the review process.