Famous investor brains, reformatted for AI agents

What it does Investor Skills is a public library of markdown skill packages that translate famous investor judgment into structured, agent-readable formats. Each skill—covering figures from Buffett to Soros to Livermore—combines YAML-encoded rules for signals, filters, sizing, and risk with Markdown prose explaining philosophy and guardrails. The repo also bundles curated reading lists, podcasts, and YouTube references to ground each model in primary sources.

The interesting bit Rather than shipping a trading engine, the project treats investor mental models as portable infrastructure: plain-text SKILL.md files that drop into Claude, Cursor, Codex, or the project’s own Questflow harness. It functions as a schema layer for financial agent behavior—part knowledge base, part prompt framework—where the product is structured judgment, not executable code.

Key highlights

• Covers diverse styles: value, macro, quant, trend following, crypto, and special situations
• Each skill packages YAML rules with human-readable philosophy and behavioral guardrails
• Designed to be portable across AI agent platforms, not locked to Questflow
• Includes templates for creating new investor skill packages from scratch
• Bundles curated media references (podcasts, YouTube, essays) for every skill

Caveats

• The free open-source skills are explicitly positioned as study and research material; execution-grade, trade-labeled skills are kept as closed-source exclusives inside Questflow
• The repository is almost entirely markdown templates and documentation rather than executable software

Verdict Worth browsing if you are building financial AI agents or want to study how great investors structure decisions. Skip it if you are looking for a backtested trading engine or traditional quantitative library.