Claude Code and Agentic Coding

This course contains the use of artificial intelligence

AI coding agents can read a codebase, write code, run tests and open pull requests. This course teaches you to direct them well with Claude Code, Anthropic's agentic coding tool, from first install to production automation.

You build TaskPilot, a full-stack task manager with Next.js, TypeScript, Prisma and SQLite, almost entirely by directing Claude Code. Along the way you learn the workflow professional teams use: explore the codebase, plan in Plan Mode, implement with tests, and commit with clear messages. You learn to write prompts the agent can verify, use slash commands, reference files and screenshots, and undo safely with checkpoints and git.

Next you master context engineering. You write a strong CLAUDEmd, manage memory at user, project and folder level, keep the context window clean with /context, /compact and /clear, and use spec driven development for larger features.

Then you extend Claude Code. You create custom slash commands, Agent Skills, subagents and hooks, share settings and permission rules with your team, and package everything as a plugin. You connect Claude to GitHub, a browser and your database with the Model Context Protocol, and build your own MCP server in Python.

After that you automate. You run Claude Code in headless mode and in GitHub Actions, set up automated pull request review, run parallel agents with git worktrees, and build a custom release notes agent with the Claude Agent SDK.

Finally you secure your workflow, control cost, set up a team, and Dockerize and deploy TaskPilot with continuous integration.

Every lecture includes captions and a transcript, and every command is real, so you can reuse it in your own projects.