zhangxuefeng-skill is a decision-making framework that encodes the thinking patterns of a career and education counselor into a structured knowledge system for AI agents.
The tool addresses the problem of making major life decisions—college major selection, graduate school choices, and career planning—by distilling a specific person's decision-making approach into reusable mental models. Rather than collecting quotes, it extracts five core cognitive frameworks including the "social sieve theory" (how credentials and assets filter life outcomes), "employment backtracking method" (analyzing where graduates actually end up), "layer realism" (how family resources constrain strategy), and "irreplaceability testing" (what skills survive technological disruption). These models are implemented as executable reasoning patterns that can be applied to new questions, not merely retrieved as advice.
The project is designed for developers building AI agents who want to add specialized decision-making capabilities. It works within the Agent Skills protocol, making it compatible with multiple AI agent runtimes. Someone considering adoption should understand this is a reference implementation showing how to encode a specific person's reasoning into a skill module—useful if you need to capture domain expertise or a particular decision-making philosophy into a format that AI systems can execute. The tool is not a general career counselor but rather a way to make one specific counselor's frameworks available to agents. It can be installed into compatible runtimes automatically or used by copying its content into conversations directly.
Development activity shows a focused effort on research depth and framework extraction. The project is grounded in documented sources including five published works, extensive media interviews, direct quotes, and a complete timeline of key decisions. The implementation demonstrates attention to making frameworks executable rather than merely descriptive, with example conversations showing how the mental models apply to concrete scenarios. The skill is designed to work across multiple runtime environments, indicating concern for portability and integration rather than building a standalone application.
