ai-berkshire
An open-source framework for value investors that helps produce structured company research and draft investment theses with Claude Code / Codex using Buffett-style methods.
Tool overview
In adoption terms, ai-berkshire currently looks more like a high-attention open-source investment research framework than a broadly validated investment product. GitHub stars, Trending visibility, and strong reposts on X are proof of attention, not proof of effectiveness. The stronger evidence for actual usefulness comes mainly from the official repository’s workflow description and a small number of long-form Zhihu explainers. The current sample still lacks broad hands-on testing, long-term reviews, and many reproducible evaluations, so capability judgments should stay conservative.
In practice, it does not automatically make money for you, and it is not a quant trading engine, backtesting platform, or broker execution tool. A more accurate analogy is a reusable research workflow template for value investing. Based on the repository description, it breaks down the perspectives of Buffett, Munger, Duan Yongping, and Li Lu into multiple agents that generate draft analysis on fundamentals, moat, risks, and valuation in parallel, then cross-check those views to help users organize research materials and decision logic faster.