What does this skill do, and when should you use it?
This is a knowledge skill for AI agents (equally readable by humans) covering investment methodology. It distills the approaches of 15 top fund managers — Buffett, Dalio, Simons, Soros, Howard Marks, Peter Lynch, Duan Yongping, Liang Wenfeng and others — into 15 standalone profiles covering selection logic, position sizing, risk control, and exit criteria. It ships with SEC EDGAR CIK numbers for 7 institutions so agents can look up quarterly 13F holdings. All sources are verifiable public materials: shareholder letters, books, memos, interviews, and 13F filings. The project also synthesizes 5 principles common to all 15 masters, ready for report generation.
Reads the 15 master profiles in the masters/ directory (buffett.md, bridgewater.md, etc.) and responds to natural-language prompts per SKILL.md triggers: distilling one master's full methodology (core principles, position management, risk control, latest holdings, takeaways); comparing two masters side by side on risk or cycles; querying quarterly 13F holdings via SEC EDGAR (free, public, updated quarterly with ~45-day lag); and generating research reports or articles built on the 5 common principles. Outputs are expected to carry source citations.
- A retail investor wants an AI to analyze a stock or today's market through a master's framework ('How would Soros view today's market?')
- An investment researcher needs to contrast schools of risk management ('Compare Dalio vs Marks on risk')
- Someone tracking institutional money wants to know what Bridgewater or Berkshire bought last quarter
- A finance content writer wants sourced articles on cycle theory or the 5 common principles
- A systematic trader wants to map discretionary masters' heuristics into quantifiable rules
- A newcomer learning investing wants to study masters one by one by style: value, macro, quant, technical
How do you install this skill?
- 13F and memo updates rely on overseas sites (SEC EDGAR, oaktree.com) that may be unreachable from mainland China; prepare network access or mirrors yourself.
- SEC EDGAR rate-limits automated requests; high-frequency AI fetching may be rejected and the skill offers no failure-handling guidance.
- Content is methodology distillation, not investment advice; attributions are source-level only and quotes were not verified line by line.
- No versioning or changelog; verify profile content and freshness before use.
- Publisher identity is unverified; attribution rests on self-declared AlphaGBM links.
- Network access
- Local filesystem
Clone the repo as a skill folder: for Claude Code run git clone https://github.com/AlphaGBM/investment-masters.git .claude/skills/investment-masters; for Cursor run git clone https://github.com/AlphaGBM/investment-masters.git .cursor/skills/investment-masters. The repo documents no other installation paths.
git clone --depth 1 https://github.com/AlphaGBM/investment-masters.git ~/.claude/skills/investment-mastersGenerated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- Compare Dalio and Marks on risk management
- What did Bridgewater buy last quarter?
- Write a research report on the 5 common principles
After installing, prompt naturally: 'Distill Buffett's investment methodology', 'Compare Dalio and Marks on risk management', 'What did Bridgewater buy last quarter?', or 'Write a research report on the 5 common principles'. For 13F queries the skill uses its built-in CIK table to access SEC EDGAR.
What are this skill's strengths and limitations?
- Every claim must carry a source — not 'AI roleplay as Buffett' style quote-dropping
- Covers the full chain: selection, sizing, risk, exit — not just famous one-liners
- 13F data comes straight from SEC EDGAR: free, public, independently verifiable
- 15 standalone profiles, MIT licensed, extensible and bilingual (EN + CN)
- Well-balanced sample spanning West and China, including Duan Yongping, Zhang Lei, and Liang Wenfeng
- 13F lags ~45 days and only shows US long equity — no shorts or off-balance-sheet positions
- Renaissance Medallion and AQR's proprietary funds don't file 13F, so their holdings can't be tracked
- It's a pure text methodology library with no backtest scripts or code; 'mapping to systematic rules' relies on model reasoning, not validated strategies
- No test suite or third-party verification; profile quality depends on the maintainers' curation
- Official install instructions cover only Claude Code and Cursor; other clients require improvisation
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Investment Masters this page | 50 · Use with care | ★ 634 | 1mo ago | MIT |
| StockValuation.io Local Valuation Workflow | 58 · Recommended | ★ 33 | 1mo ago | Apache-2.0 |
| Deep Company Series (看懂XX) | 49 · Use with care | ★ 17k | 14d ago | MIT |
| EdgarTools — SEC EDGAR Filing Analysis Skill | 67 · Recommended | ★ 2.8k | 11d ago | MIT |
| Global Stock Data — Full-Stack US & HK Market Data Skill | 62 · Recommended | ★ 1.7k | 14d ago | Apache-2.0 |
The README self-positions against 'AI roleplay tools' (surface quotes from training data) and 'generic finance bots' (price alerts and news feeds), distinguishing itself via primary sources, full methodology chains, and 13F tracking. No specific named competitors are given.
How did FollowSkills review this skill?
Pure knowledge skill: no scripts, no permissions, no external side effects; 13F sourcing explicitly points to free public SEC EDGAR data with CIKs, so data flow is transparent. Deducted for: master quotes attributed only to book/letter titles without per-claim verifiable links; MIT license year 2026 is questionable; publisher unverified with attribution resting on self-declared links.
Instruction-only text with no scripts or tests; static review cannot verify key paths. 13F tracking depends on runtime AI access to EDGAR with no error handling or failure feedback. Deducted for: no tests, no abnormal-input behavior, and referenced masters/*.md profile files not visible in evidence.
Clear triggers and scenarios (methodology distillation, comparison, 13F lookup), with a stated boundary that some masters are not in 13F; reasonably friendly to Chinese users (Duan Yongping, Liang Wenfeng, Chinese sources). Deducted for: reliance on overseas sources (SEC EDGAR, oaktree.com) that may be unreachable from mainland China with no fallback disclosed; no non-fit disclaimer (e.g., not investment advice).
Well-organized docs, install instructions, example queries, update schedule, license, and related links. Deducted for: no versioning or changelog, maintenance responsibility only implied (PRs welcome), and README references a master-profile template 'in SKILL.md' that is not present in SKILL.md.
Knowledge-distillation outputs are directly generatable by AI; example queries are concrete and the 5 common principles carry substantive content. Deducted for: static review cannot confirm output completeness or direct usability; 13F freshness claims unverified; marginal value over manual research unevidenced.
Each methodology points to primary sources (books, letters, interviews) and CIKs are independently checkable — limited reproducibility. Deducted for: no tests, no verified representative outputs, citations lacking specific links/pages, no fact/inference separation, thin coverage.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →