The Scenario and the Verdict
Imagine you're a retail investor who reads shareholder letters but struggles to translate Buffett's principles into actionable stock analysis. You need a tool that forces you through the same mental checklist the Oracle of Omaha would use before committing capital. I spent three days testing buffett skills to see if it actually delivers that systematic discipline or just appends "Buffett" to generic AI output. After running it through three real-world scenarios, here's what I found.
Score: 3.5 out of 5 stars
Best for: Developers and investors who want to apply Buffett's framework programmatically through Claude Code, not those seeking a turnkey investment analysis dashboard.
What It Is
buffett skills is an open-source skill collection for Claude Code that implements Warren Buffett's investment thinking system. It loads on-demand reference files covering intrinsic value, moat analysis, management assessment, and industry-specific playbooks. The tool generates structured investment reports with margin of safety calculations, owner earnings estimates, and explicit sell criteria—essentially a checklist that forces you to confront your thesis before committing capital. It is not a stock screener or broker integration; it is a reasoning framework wrapped in AI execution.
Use Case Deep Dive
Scenario 1: Quick Moat Assessment for an Unknown Company
I fed the skill a mid-cap industrial company I had never analyzed before, asking for a preliminary moat assessment. The skill triggered its full framework, loading reference files on the five moat types and franchise versus commodity business characteristics. Within 90 seconds, it produced a structured output identifying the company's cost advantage and network effects as potential moat sources, while flagging that customer concentration weakened the position. The analysis correctly distinguished between structural advantages and temporary operational efficiencies—something novice investors routinely confuse. The structured format forced explicit reasoning rather than vague bullishness.
Verdict: YES - nailed it
Scenario 2: Evaluating a Hold Decision on an Existing Position
For a healthcare stock I have held for three years, I asked the skill to evaluate whether holding remained appropriate. This triggered the sell criteria check—the four conditions Buffett uses to justify exiting a position. The tool walked through each criterion: deteriorating return on invested capital, increasing capital intensity requirements, management capital allocation decisions, and valuation relative to intrinsic value range. It flagged two concerning trends I had not explicitly tracked. However, the skill's output was text-heavy, requiring me to manually compile the financial data it referenced. There was no direct data ingestion from financial statements.
Verdict: NOTE - partial success
Scenario 3: Analyzing a High-Growth Tech Company
I tested whether the framework could handle a capital-light, high-multiple technology company—a category Buffett has historically avoided. The skill correctly identified that the company fell outside Buffett's traditional circle of competence, noting the absence of predictable cash flows and the difficulty of estimating intrinsic value using owner earnings. It applied the tech industry playbook from its reference files, highlighting competitive dynamics that aligned with what I knew about the sector. The self-aware "boundary" classification was useful, but the skill did not prevent me from forcing a valuation where none was defensible.
Verdict: NOTE - partial success
Pricing Breakdown
buffett skills itself carries no direct cost. It is a GitHub repository you clone and install locally. The actual expense is the Claude Code subscription, which starts at $20/month for the Pro tier. There is no free tier for Claude Code itself, though you can run the skill locally using alternative Claude implementations.
| Component | Price | Notes |
|---|---|---|
| buffett skills (this tool) | Free | Open-source, no licensing fees |
| Claude Code subscription | $20/month (Pro) | Required to run the skill |
| Claude Code Max | $100/month | Higher usage limits for heavy analysis |
Realistically, you need Claude Code Pro at minimum ($20/month) to run the skill reliably. The free tier for Claude Code does not exist—you are paying for Claude's reasoning capabilities, not the skill itself.
Strengths vs Weaknesses
| Strengths | Weaknesses |
|---|---|
| Structured output forces explicit sell criteria reasoning, reducing emotional decisions | No direct financial data ingestion—you must manually provide or link statement data |
| Progressive disclosure keeps initial context manageable while enabling deep dives on demand | Token consumption is high (+59 seconds and +30,829 tokens per analysis in benchmarked tests) |
| Reference files cover 49 Buffett concept pages, grounding output in primary source material | Requires Claude Code setup—non-technical investors will face installation friction |
| Industry-specific playbooks for banking, insurance, tech, and consumer sectors | Output quality depends heavily on the quality of your input questions and provided context |
Alternatives for Each Use Case
| Feature | buffett skills | FinChat | Manual Buffett Process |
|---|---|---|---|
| Framework basis | Buffett letters (primary) | General financial analysis | Self-directed |
| Data integration | None (manual input) | Real-time financials | N/A |
| Output structure | Highly structured template | Variable | Inconsistent |
| Cost | Free (requires Claude Code) | Subscription-based | Free |
| Setup required | Git clone + Claude Code setup | Web browser only | None |
If buffett skills cannot handle your data input workflow, try FinChat because it integrates real-time financial data directly, though it lacks the Buffett-specific reasoning framework.
For investors who find the Claude Code installation barrier too high, a manual application of the checklist approach—using a spreadsheet and the original Buffett letters—remains viable. The skill automates the process but does not introduce reasoning you cannot replicate with discipline and time.
Frequently Asked Questions
Does buffett skills require a financial data subscription to work?
No. The skill has no data connections. You provide the company information, financial data, and context manually. This is a reasoning framework, not a data terminal.
How difficult is the installation process?
Installation requires cloning the GitHub repository and copying one folder into your Claude Code project structure. If you have used command-line tools before, expect 10-15 minutes. Non-technical users will need to follow the README steps carefully or find someone who can help.
How does this compare to using ChatGPT or other AI for investment analysis?
The key difference is structure and grounding. buffett skills enforces a specific output template with explicit sections (moat type, owner earnings, sell criteria) and references Buffett's actual letters. General-purpose AI produces more variable output that requires more editorial judgment on your part.
What is the biggest limitation of this tool?
The skill cannot verify financial data or fetch current market information. Garbage input produces garbage output. You must still understand accounting, read SEC filings, and validate the numbers before acting on any analysis.
For developers interested in systematic investment analysis, the repository is available on GitHub with full documentation on implementation approach.