Code, prompts and files from the hands-on tutorials on the skill.md YouTube channel.
One folder per episode, numbered in the order they come out. The episode number is shown in the bottom-left of every video and on its end card (→ ep001), and the video description links straight to its folder.
| Ep |
Date |
Video |
Folder |
| 001 |
2026-10-22 |
Ollama Tutorial: A Free Local AI Built This Game in 3 Prompts |
ep001-ollama-local-ai-game/ |
| 002 |
2026-10-23 |
MCP Server Tutorial (Python + Ollama): My Local AI Stopped Guessing |
ep002-mcp-server-ollama/ |
| 003 |
2026-10-27 |
How to Make Money With AI in 2026: 7 Steps (Real Run, Real Mistakes) |
ep003-make-money-with-ai/ |
| 004 |
2026-10-26 |
Claude Mythos 5 Went Rogue in a Safety Test. It Faked a Human |
ep004-ai-agent-network-sandbox/ |
| 005 |
2026-10-29 |
Claude Code Skills Tutorial: Your First SKILL.md, Free With Ollama |
ep005-claude-code-skills-ollama/ |
| 006 |
2026-10-31 |
Claude Code Hooks Tutorial: The AI Found a Way Around My Rule |
ep006-claude-code-hooks/ |
| 007 |
2026-10-29 |
Downloading the Claude Desktop App? Fake Ones on GitHub Hide Malware |
ep007-check-ai-app-downloads/ |
| 008 |
2026-10-30 |
Ollama RAG Tutorial: Chat With Your PDFs Locally (No LangChain) |
ep008-local-rag-ollama/ |
| 009 |
2026-11-02 |
AI News Today: Claude Fooled Anthropic's Safety Monitor |
ep009-daily-news-checks/ |
| 010 |
2026-11-03 |
Build an AI Agent in Python From Scratch: Right Math, Wrong Answer |
ep010-python-ai-agent-ollama/ |
| 011 |
2026-11-04 |
Claude Code Subagents vs Agent Teams: Same Answer, 9x the Tokens |
ep011-subagents-vs-agent-teams/ |
| 012 |
2026-11-05 |
Postman Claude Code Plugin Installed Itself. Did You Say Yes? |
ep012-postman-agent-plugin-check/ |
| 013 |
2026-11-06 |
OpenAI Linux Kernel Exploits: AI Built 948 Root Hacks for $1.2M |
ep013-linux-kernel-patch-check/ |
| 014 |
2026-11-07 |
NVIDIA Auto Ontology: Stops AI Agents Guessing Your Data (Mostly) |
ep014-auto-ontology-duplicate-columns/ |
| 015 |
2026-11-07 |
AI News Today: DeepMind AI Agents Found 2 Holes in Old RSA Encryption |
ep015-openssh-rsa-key-check/ |
| 016 |
2026-11-08 |
AI Automation for Small Business: My AI Said Yes to a Scam |
ep016-ai-inbox-automation/ |
| 017 |
2026-11-08 |
Claude Code Security: "Don't Ask Again" Leaked 2,539 Secrets |
ep017-claude-code-secrets-check/ |
| 018 |
2026-11-09 |
Vercel Agent Browser: 47K Stars, Every Safety Switch Off by Default |
ep018-agent-browser-safe-start/ |
| 019 |
2026-11-09 |
Nvidia CEO Jensen Huang Under Investigation by His Own Board (story 3: DMARC check) |
ep019-dmarc-check/ |
| 020 |
2026-11-10 |
Whisper AI Transcription: Free & Offline, But It Wrote "Cloud Code" |
ep020-whisper-transcription/ |
| 021 |
2026-11-10 |
Claude Fable 5.2: 1,000 AI Agents Formalized Fermat in 30 Hours (story 1: API call) |
ep021-claude-fable-5-2-api/ |
| 022 |
2026-11-14 |
OpenAI Codex Security: No Cost Cap by Default, 96-Hour Deep Scans |
ep022-codex-security-safe-start/ |
| 023 |
2026-11-15 |
llama.cpp Tutorial: I Asked for a Poem, It Asked to Run JavaScript |
ep023-llamacpp-local-server/ |
| 024 |
2026-11-16 |
Chrome DevTools MCP: It Found Our Bug. Then We Changed 4 Defaults |
ep024-chrome-devtools-mcp-real-test/ |
Each folder has its own README: what the video shows, every command and prompt in order, the files the AI produced (unedited), and the licenses of the tools and models used.
Everything in a folder is what we actually ran and what the AI actually produced. Check each tool's and model's license before commercial use.
The hands-on, show-every-step part of these tutorials is inspired by FreeAIVault.