
Discover quick information about cloud code, harnesses, and core AI concepts like LLMs, tokens, APIs, and function calling, and learn when to watch the next videos for deeper understanding.
Discover how Cloud Code wraps a language model in a harness that provides tools, context, permissions, memory, and an action loop for completing tasks.
Explore AI basics, including LLMs, tokens, and API usage, to understand context windows, token pricing, and function calling across skills, plugins, and CLIs for developers.
Learn to integrate Cloud Code inside an IDE, focusing on VS Code and its forks like Cursor and Antigravity, and set up Cloud Code in the terminal.
Expand your AI agent with the model context protocol and Zapier MCP server, connect to Gmail, Calendar, and 8000 apps, and extend capabilities with skills, plugins, and CLIs.
Explore how to manage permissions and modes in Cloud Code, including allow, deny, and ask lists, plus default, auto, plan, and bypass modes, with practical examples.
Design a soul.md to give your ai assistant a defined personality, tone, bravery, and humor, with clear boundaries and default levels, and integrate it into cloud code by updating cloud.md.
Create and manage custom slash commands in the .cloud/commands folder to automate prompts like macro and push, fetch data from federalreserve.gov, and keep prompts up to date.
Build a personal wiki and second brain with Obsidian and Claude Code, using markdown files, backlinks, and VS Code to ingest, connect, and query knowledge via a coding agent.
Explore the latest model lineup in cloud, desktop, and terminal. Use Opus for most tasks, reserve Fable 5 for hard problems, and avoid Sonnet 5.
Recap the cloud code setup, highlight native installation for flexibility, and summarize context engineering, prompts, and plugins like Zapier and Notebook LM to build a personal agent.
Test and refine a Next.js marketing site for AI Arnie by starting the dev server with npm run dev, creating a Cloud.md for Cloud.code, and checking navigation.
Host your webpage on vercel by connecting github, importing your project, and deploying with the next.js preset. See how vercel and github sync automatically to keep updates live.
Use the slash btw /btw to ask quick side questions while cloud code runs, without interrupting the session, and optionally open a new terminal.
Explore cloud design by Entropic Labs to turn static mockups into interactive prototypes, pitch decks, and marketing collateral, exportable to PPTX or Canva, with fully branded design systems.
Explore advanced cloud code concepts for building and running agents, including sub-agents and parallel execution, using the cloud agent SDK in Python or TypeScript to support NADN workflows.
Learn to create and run sub-agents in parallel with Cloud Code, use Context 7 for documentation, and build dedicated backend, frontend, and testing agents that have separate context windows.
Learn how auto research builds a self-improving ai that autonomously edits code, runs training, and iteratively improves skills like an OpenClaw documentation search, inspired by Andrej Karpathy’s open source project.
Discover how to set up Claude Code desktop and enable computer use to let the app control your keyboard and mouse across Mac and Windows.
Get a concise overview of every slash command in Cloud Code, from status line and loops to advisor, branches, and plan modes, with tips on token use and debugging.
Master NADN workflows with Cloud Code, connect to an MCP server, and manage permissions, API keys, and skills to build and monitor JSON-based automations.
Secure your cloud code by keeping API keys in .dev files, never pushing secrets to GitHub, and rotating them regularly to guard against supply chain attacks and MCP server risks.
Become a cloud code pro by building AI agents with MCPs, skills, CLIs, and tools, integrating NADN and the Agent SDK to deploy workflows.
Claude Code is not just another AI coding assistant.
It is a new way to work with your computer: with your files, your terminal, your codebase, your tools, your automations, and your projects.
In this course, we turn Claude Code from “interesting AI tool” into a practical AI work environment.
The course is built around 3 main layers:
1. Claude Code Foundations
Installation, setup, IDE workflows, terminal basics, project structure, CLAUDE md, the .claude folder, modes, permissions, slash commands, custom commands, and the core workflow of working with Claude Code in real projects.
2. Practical AI Projects
Instead of only learning features one by one, we build real systems: a personal AI assistant, an Obsidian LLM wiki, a website from idea to deployment, n8n automation workflows, self-improving AI with autoresearch, and agents with the Claude Agent SDK.
3. Advanced Agentic Workflows
We go deeper into context engineering, MCP, skills, plugins, CLIs, NotebookLM, RAG, Telegram control, cron jobs, loops, hooks, subagents, parallel agents, Context7, Git, GitHub, Vercel deployment, Claude Desktop, Computer Use, and security.
The goal is simple:
By the end of the course, Claude Code should not feel like a mysterious AI tool anymore.
It should feel like a practical system you can use to build, automate, research, test, deploy, and manage real projects.
The course includes 6 major hands-on projects:
1. Personal AI Assistant
Build a Claude Code assistant connected to Telegram, channels, plugins, skills, MCP tools, loops, scheduled tasks, and custom instructions. Similar in spirit to OpenClaw, but built around Claude Code.
2. Obsidian LLM Wiki
Create a second brain system with Obsidian and Claude Code for memory, documentation, RAG-style knowledge workflows, and personal knowledge management inspired by ideas from Andrej Karpathy.
3. Website Project with Claude Design
Build a website from the first idea to frontend design, testing, GitHub version control, Claude Design, and final deployment on Vercel.
4. n8n Automation System
Use Claude Code to create, improve, and manage n8n workflows for AI automation, workflow building, and practical automation systems.
5. Self-Improving AI with Autoresearch
Build an autoresearch workflow where Claude Code can research, improve, iterate, and work more autonomously on complex tasks.
6. AI Agents with Claude Agent SDK
Build agents in Python with Anthropic’s Claude Agent SDK, including a customer support agent and bonus workflows with Pydantic-based agents.
This is not a course about random prompting tricks.
It is a project-based Claude Code course for people who want to understand how AI coding tools, AI agents, automation, context engineering, MCP, RAG, GitHub, deployment, and real-world workflows fit together.
We start with the basics, including AI and API fundamentals for developers: LLMs, tokens, function calling, and the concepts behind modern AI tools.
Then we move step by step into Claude Code itself: installation, IDEs like VS Code, Cursor, and Google Antigravity, project structures, permissions, modes, context files, slash commands, custom prompts, skills, plugins, and CLI workflows.
After that, we build real systems.
You see how Claude Code can work with Telegram, Obsidian, n8n, NotebookLM, GitHub, Vercel, Claude Desktop, Computer Use, Context7, MCP servers, hooks, logs, scheduled tasks, subagents, and parallel agents.
The final section covers important security topics like tool poisoning, MCP rug pulls, permissions, AI agent vulnerabilities, and safe usage when giving AI access to files, tools, APIs, workflows, and your computer.
This course is for developers, AI builders, automation creators, n8n users, entrepreneurs, freelancers, students, content creators, and knowledge workers who want to use Claude Code in practical projects.
You do not need to be an expert developer to start.
The course begins with the basics and then moves into advanced Claude Code workflows, AI agents, automation, deployment, RAG, MCP, hooks, subagents, and real-world projects.
If you want to stop using AI only as a chatbot and start building useful AI systems with Claude Code, this course gives you the complete practical roadmap.