Claude: From First Chat to Building Agents

Claude: From First Chat to Building Agents teaches everything about Claude, the AI made by Anthropic, starting from zero and going deep enough for professionals who build with it every day. You'll begin with how language models work in plain words (tokens, the context window, thinking, why AI sometimes makes things up), the Claude model family and how to choose between Fable, Opus, Sonnet and Haiku, what the plans and the API cost, every place Claude lives, and how safety and your data are handled.

Then you'll master the Claude apps on web, desktop and phone: files and PDFs, web search and Research, artifacts (Docs, Slides, Design and small apps), projects, memory, connectors, skills, plugins, tasks that Claude carries out for you, and Claude inside Chrome, Excel, PowerPoint, Google Workspace and Slack. A module on talking to Claude well covers briefing, examples, XML tags, long documents, format and voice, checking answers and a library of prompt patterns.

A module called The bits nobody explains teaches the terminal, files and paths, git and GitHub, APIs, JSON, keys and installing Python or Node, so beginners can follow everything after it. Three modules then cover Claude Code in full: installing it, your first session, steering, the context window, CLAUDE.md and memory, permissions and the sandbox, settings, every command, keyboard shortcut and command-line flag, skills, subagents, hooks, MCP, plugins, parallel agents, cloud and remote work, scripts, CI and code review.

Finally you'll build with the Claude API (messages, thinking, effort, structured outputs, images and PDFs, citations, tool use, Anthropic's ready-made tools, caching and batches) and build agents with the Claude Agent SDK and Claude Managed Agents, including your own MCP servers and skills, evals and safe deployment. The course ends with a capstone project and a method for staying current as Claude changes. 64 lessons in 10 modules.