The Claude Code Guide Series
I’ve put together a 7-part guide series on Claude Code, covering everything from installation through to coordinating agent teams. Each guide builds on the previous one, but they all stand on their own if you want to jump around. The full series is in the guides section.
Installing Claude Code — Getting set up from scratch. Authentication options, updating, and the different ways to connect to Claude (direct API, Anthropic Max subscription, or third-party providers like Bedrock and Vertex).
Using the Claude Code TUI — The terminal interface itself. Modes (normal, plan, compact), keyboard shortcuts, slash commands, and how the permission system works. This is the stuff you’ll use every single session.
CLAUDE.md and Settings — Project context and configuration. How to write effective CLAUDE.md files, the settings hierarchy, and how to set up permissions properly. This is where most people start getting significantly better results.
Skills, Agents, and MCP Servers — Extending Claude Code’s capabilities. Skills for reusable knowledge and workflows, sub-agents for delegation, and MCP servers for external integrations. Also covers why wrapping MCP servers in CLI scripts is often a better approach for context efficiency.
Hooks — Lifecycle events that let you run code when specific things happen. Blocking dangerous commands, validating tool use, injecting context, and enforcing project rules automatically.
Plugins and Marketplaces — How skills, agents, hooks, and MCP servers get packaged and distributed. Installing plugins from marketplaces, managing them across projects, and enterprise distribution.
Agent Teams — Running multiple Claude Code instances in parallel, coordinated through shared tasks and inter-agent messaging. This is where things get genuinely powerful for larger work.