Overview

AI Berkshire is a cutting-edge value investing research framework designed for the AI era, built upon Claude Code. It systematically integrates the methodologies of four renowned value investing masters—Warren Buffett, Charlie Munger, Duan Yongping, and Li Lu—and leverages multi-agent adversarial analysis to achieve professional-grade investment research. This framework transforms a single individual with Claude into a potent investment research team, significantly enhancing research depth and efficiency.

Key Features

• AI-Powered Expert Methodologies: Employs structured frameworks derived from Buffett, Munger, Duan Yongping, and Li Lu, providing a multi-dimensional analysis of investment opportunities.
• Multi-Agent Adversarial Research: Utilizes parallel AI agents to research a single company from different perspectives, fostering genuine conflict and tension in analysis to uncover blind spots and avoid single-point failures.
• Decision-Oriented Outputs: Moves beyond general analysis to provide clear, actionable investment conclusions: "Pass," "Fail," or "Gray Area," complete with price ranges and stratified recommendations. Includes a "Mirror Test" for concise decision justification.
• Structured Bias Mitigation: Integrates mechanisms like information richness grading, Munger-style contrarian testing, rapid rejection checklists, and anti-consensus checks to ensure robustness and prevent misleading AI outputs.
• Financial Rigor & Precision: Implements strict data validation, using Python's decimal.Decimal for all calculations to ensure absolute precision, and cross-verifying critical financial data from at least two independent sources.
• Reproducible Research Workflow: Guarantees consistent output format and depth for identical inputs, enabling easy comparison of companies, tracking of changes over time, and alignment of research across team members.
• Comprehensive Skill Set: Offers a suite of 16 specialized "Skills" for various investment research needs, from in-depth company analysis and earnings reviews to industry screening, portfolio management, and even private company research.

Typical Use Cases

• In-depth Company Analysis: Conduct comprehensive, multi-dimensional research on publicly traded companies using the /investment-research and /investment-team skills, simulating a full investment research department.
• Rapid Stock Screening: Quickly filter potential investments using the /investment-checklist (Buffett's criteria) or /quality-screen (7 hard metrics) for efficient initial evaluation.
• Industry & Sector Deep Dives: Gain a panoramic view of an industry's investment landscape with /industry-research or systematically narrow down opportunities with the /industry-funnel to identify top-tier companies.
• Financial Statement Mastery: Analyze earnings reports with unparalleled depth using /earnings-review and /earnings-team, focusing solely on primary source data.
• Portfolio Management & Monitoring: Optimize and track existing holdings with /portfolio-review and maintain discipline with the /thesis-tracker for ongoing research validation.
• Unlisted Company Evaluation: Research private companies and venture-stage investments where data is scarce using the specialized /private-company-research framework.
• Quick Event Attribution: Understand the drivers behind stock price movements in minutes with the /news-pulse skill, distinguishing between value events and mere market noise.
• Mastering Investment Philosophy: Learn and apply the thinking patterns of investing legends through specific skills like /dyp-ask (Duan Yongping's questioning style).