Summary

This project addresses the generic nature of AI-generated advice by distilling the specific mindset, decision-making framework, and communication style of creator @liyupi into an AI skill. It targets developers, job seekers, and entrepreneurs seeking practical, experience-driven insights rather than textbook answers, offering a distinct and relatable AI persona.

How It Works

The skill employs a systematic "knowledge distillation" process, embedding @liyupi's "mindset operating system" and concrete "decision rules" (e.g., "have a fallback before aiming higher," "learn what the market needs"). It goes beyond superficial tone mimicry to replicate a specific problem-solving approach and communication style, characterized by directness, self-deprecating humor, and structured responses. This allows AI assistants to provide actionable, personalized advice grounded in real-world experience.

Quick Start & Requirements

Installation involves cloning the repository into specific directories for supported AI platforms:

• Claude Code: mkdir -p .claude/skills && git clone https://github.com/liyupi/yupi-skill.git .claude/skills/yupi-skill (local) or ~/.claude/skills/yupi-skill (global).
• Cursor: mkdir -p .cursor/skills && git clone https://github.com/liyupi/yupi-skill.git .cursor/skills/yupi-skill.
• OpenClaw: git clone https://github.com/liyupi/yupi-skill.git ~/.openclaw/workspace/skills/yupi-skill. Requires Git and an AI assistant compatible with the AgentSkills standard. No other non-default prerequisites are listed.

Highlighted Details

• Distilled Mindset: Encapsulates 7 core mental models, including "Direction Judgment + Go All Out" and "MVP Shell + Core Uncompromised."
• Decision Rules: Implements practical heuristics such as "Have a fallback before aiming higher" and "Verbal promises don't count."
• Expression Style: Features "conclusion first," self-deprecating humor, numbered points, and summary closings.
• Integrated Knowledge Sources: Leverages specific domains like codefather.cn, ai.codefather.cn, mianshiya.com, laoyujianli.com, dogyupi.com, and GitHub for up-to-date information.
• Systematic Distillation: The creation process includes material collection, persona analysis, deep probing, and self-validation against real-world examples.

Maintenance & Community

The project is maintained by @liyupi (程序员鱼皮), a prominent figure known for AI and programming content, formerly a Tencent engineer and current CEO. Community engagement is primarily through platforms like Bilibili, GitHub, and dedicated navigation sites (codefather.cn, ai.codefather.cn, etc.).

Licensing & Compatibility

The project is released under the MIT License, permitting broad use, modification, and distribution, including for commercial purposes and integration into closed-source applications.

Limitations & Caveats

The skill's knowledge base is current as of April 2026, requiring web search for the latest information. It cannot replicate non-verbal communication cues from video content. The AI will explicitly state when information is not publicly disclosed. For complex technical queries, it relies on web search before applying the persona, rather than generating novel technical insights independently. It is dependent on AI platforms supporting the AgentSkills standard.