I Tried /teach and 10x'd My Ability To Learn

I Tried /teach and 10x'd My Ability To Learn

What is the /teach Skill and How Does It Work?

Introduction to the /teach Skill

  • The speaker introduces a lesson created by an AI agent using the /teach skill developed by Matt Pocock, highlighting its effectiveness in learning.
  • The /teach skill transforms Claude Code Agent into a personal tutor capable of teaching various subjects, including coding and open-source models.

Key Features of the /teach Skill

  • The video will cover how to install the skill, examples of lessons it creates, and the core idea that makes it effective.
  • Traditional AI interactions often provide inconsistent answers; this skill maintains context and remembers user progress, unlike typical substitute teachers.

Stateful vs. Stateless Learning

  • Unlike stateless tools that forget user history, the /teach skill is stateful, tracking what users have learned and where they struggle.
  • This feature allows for personalized lesson progression based on previous questions and understanding.

How Does Learning Progress with /teach?

Lesson Structure and Tracking

  • The system logs learning records, including questions asked and areas where users get stuck, creating tailored lesson plans.
  • A mission statement is generated to clarify learning goals; for example, the speaker's motivation stems from dissatisfaction with Claude Fable.

Glossary Creation

  • As lessons progress, a glossary is built automatically to define terms encountered during learning sessions.

Examples of Lessons Created by /teach

First Lesson Overview

  • The first lesson focuses on open weight models; it provides installation commands for Ollama while explaining their relevance based on hardware capabilities.

Interactive Learning Experience

  • Each lesson includes checklists for tasks completed and self-check questions to reinforce understanding through action-based learning.

The Philosophy Behind Effective Teaching

Knowledge Acquisition Process

  • Developed from ten years of teaching experience, the philosophy behind /teach emphasizes three levels: knowledge (trusted sources), skills (practical application), and wisdom (community engagement).

Community Engagement Importance

  • Users are encouraged to connect with communities related to their topics of interest as part of their learning journey.

Personalized Learning Experience

Zone of Proximal Development Concept

  • Lessons are designed at an appropriate challenge level—neither too easy nor too difficult—ensuring engagement without frustration.

Transitioning from Tutor to Community Member

  • Once users demonstrate mastery in a subject area, they are guided towards community resources rather than remaining dependent on AI assistance.

Setting Up the /teach Skill

Installation Process

  • To set up the skill, users can access a GitHub repository link provided in the video description for installation instructions.

Running Lessons

  • After installation, users can initiate lessons by specifying topics they wish to learn about within their workspace setup.

Conclusion: Emphasizing Active Learning

Importance of Personal Effort

  • While AI can assist in research and organization, true understanding requires active participation in learning processes.

Additional Resources

  • A free resource guide is available for viewers who want further assistance with installation commands or prompts related to using this tool effectively.
Video description

Work with me👩‍💻 https://deeprooted.io Free Resources✍ https://www.skool.com/ai-automation-network-2970/about Make Money With AI 🤑 https://www.skool.com/ai-automation-society-plus/about?ref=6d792349a24746cba3127867f850d497 There's a free Claude Code skill called /teach that turns your agent into a personal tutor, and it can teach you literally anything. Code, a new framework, even how open source models actually work, which is exactly what I've been using it for. It was built by Matt Pocock, the same guy behind the grill-me skills, and honestly it might be the most useful skill I've installed all year. The thing that makes it different is that it's stateful. Most AI answers are a one-off, like a substitute teacher who has never met you and forgets you the second the bell rings. /teach saves a whole teaching workspace to your file system, the mission, the resources, your numbered lessons, your learning records, a glossary, so it always knows exactly where you left off and builds your next lesson on purpose. In this one I walk through the real lessons it built me on open source models, the Knowledge to Skills to Wisdom idea that makes it click, and how to install it and start learning in about five minutes. 🛠 My Tools: HyperFrames🎬 : https://hyperframes.heygen.com Glaido🎙 : https://glaido.com 🔗 Resources Download Claude Desktop: https://claude.ai/download Claude Code: https://www.anthropic.com/claude-code /teach skill (Matt Pocock)🛠 : https://github.com/mattpocock/skills Matt Pocock's Video: https://youtu.be/s5T5oQJcJ6U?si=SnC-kHEm4AuaZyqY TIMESTAMPS: 00:00 - Intro 00:50 - Stateful vs Stateless 02:00 - Teaching Workspace Structure 03:20 - What A Lesson Looks Like 06:50 - Pocock's Philosophy 07:50 - The Learning Zone 08:30 - Setting You Free 09:05 - Installing /teach 09:30 - Full Demo 12:15 - You Still Do The Work 12:45 - Free Resources 13:05 - Outro 13:35 - Blessing #claude #tearning #teaching #mattpocock