Basics of Claude Code — Workshop

Basics of Claude Code — Workshop

A practical introduction to Claude Code for anyone looking to start using it for coding.

Pre-Requisites:

  • Claude Code account — check GitHub settings

  • Latest version of Claude Codehttps://code.claude.com/docs/en/quickstart

  • VS Code — Or any other code editor will do

  • A Codebase you want to use to practice talking to Claude Code.

  • Terminal — Mac’s Terminal or Windows Powershell will work, you can try Warp if you want an AI-aware terminal

Duration: ~2 hours

https://docs.google.com/presentation/d/1K-QU4tRBbnq5UH7Z5jlppUS8ALrbZBTXY-RyCYBnqfs/edit?usp=sharing

Welcome to Claude Code Basics! This training takes about two hours to go through, and it's designed so you can either follow along in a live session or work through it on your own at your own pace. There are two hands-on exercises built in so you can practice as you go.

Here's what we're going to cover:

  • First half: What LLMs and AI agents are, and how Claude Code works — models, modes, tools, and context

  • Second half: More advanced workflows like running multiple agents and connecting Claude Code to external services with MCP

The goal here isn't just to learn Claude Code — it's to learn how to think in terms of AI agents. Once you've got that down, you'll be able to pick up any of the other tools out there (Cursor, Copilot, Windsurf, etc.) pretty easily, because they all work on the same ideas.

Alright, let's get into it.


Workshop Sections

Work through each section in order and complete the exercises as you go!

CC - LLMs vs AI Agents

CC - Models & Effort

CC - Conversational Mode

CC - Tools

CC- Plan Mode

CC - Agent-mode

CC - Context Management

CC - History & Checkpoints

CC - Multi-Agent Workflows

CC - Model Context Protocol (MCP)


Wrapping Up

That's a wrap on Claude Code Basics! Here's a quick recap of everything we covered:

  • What LLMs are and how they work — tokens, vectors, next-token prediction

  • What AI agents are and how they use LLMs as a brain

  • The Claude Code modes — Conversational, Plan, and Auto-mode — and when to use each

  • How to choose models and manage your request budget

  • How tools work and how to control which ones Claude Code can access

  • Adding context and checkpoints

  • Running multiple agents simultaneously and using subagents for parallel work

  • Extending Claude Code's capabilities with MCP servers

Everything here applies broadly — the mental model transfers to Cursor, Copilot, Windsurf, and any other agent you pick up down the road.

Thanks for following along!

Content Index

Core Concepts

Concept

Summary

Section

Concept

Summary

Section

LLM (Large Language Model)

The core AI technology — a next-token prediction machine trained on massive amounts of text.

 

Token

The unit an LLM actually predicts — smaller than a word, a chunk of text that maps to a number.

 

Vector / Embedding

How text is converted into numbers and placed in a high-dimensional semantic space so the model understands relationships between concepts.

 

Next-token prediction

The fundamental mechanism of LLMs — given some input text, predict the most likely next token.

 

AI Agent

A software layer built on top of an LLM that gives it "hands" — the ability to read/write files, run commands, call APIs, and take real actions.

 


Claude Code Models & Configuration

Concept

Summary

Section

Concept

Summary

Section

Model Picker

How and why to select the different models

 

Effort

A toggle on some models that controls how much reasoning the model does before responding.

 

Usage

How usage is measured, how its affected by model + effort, and how to check your current usage

 


Claude Code Modes

Concept

Summary

Section

Concept

Summary

Section

Conversational Mode

Conversational-only mode — no file edits or commands. Best for brainstorming, exploring ideas, and building shared understanding before writing code.

 

Plan Mode

Optimized for creating implementation plans. Has its own tool set tuned for research and step-by-step planning. Best used with a powerful model.

 

Agentic Mode

Full-power mode — Claude Code can edit files, run terminal commands, execute tests, spin up servers, and orchestrate multi-step work.

 

Ask → Plan → Agent workflow

The recommended sequence: design in Ask, plan in Plan, implement in Agent. Produces better results than jumping straight to implementation.

 


Tools & Context

Concept

Summary

Section

Concept

Summary

Section

Tools

Built-in capabilities Claude Code can invoke (fetch, file edit, terminal, etc.). Available tools vary by mode. Viewable via the tools icon in the chat.

 

Fetch Tool

Lets Claude Code browse the web.

 

Approvals Dropdown

Controls when Claude Code asks for confirmation before taking actions: Default (Claude Code decides), Bypass (auto-approve), or Autopilot (fully autonomous).

 

Code Highlighting

Select lines in the editor to automatically pin them as context in the chat window.

 

Drag and Drop Files

Drag files or images from the file explorer directly into the chat to add them as context.

 

Context Window

Check how full Claude Code's memory is. Start a new chat when it gets too full.

 

Checkpoints

You can undo recent changes. Use git to make checkpoints so you can roll back all subsequent file changes.

 

Conversation History

Every message includes the full chat history, so Claude Code remembers everything discussed — the more you flesh out in Ask mode, the better it implements later.

 


Multi-Agent Workflows

Concept

Summary

Section

Concept

Summary

Section

Multiple Chat Windows

Open several independent Claude Code chat windows side by side — each with its own mode, model, and context — all scoped to the same project.

 

Subagents

A built-in tool (#run-subagent) in Agent/Plan mode that lets Claude Code spin up and orchestrate its own internal agents to run tasks in parallel.

 

Isolated Context Window

Each subagent runs in its own context — it only knows what the main agent passes it. Intermediate thinking is discarded after it returns its result.

 

Manual multi-agent

You orchestrate multiple chat windows yourself — good for a small number of independent tasks running in parallel.

 

Automated multi-agent

Ask Claude Code to use subagents and it handles all orchestration internally — good for coordinating everything toward a single goal.

 


Model Context Protocol (MCP)

Concept

Summary

Section

Concept

Summary

Section

MCP (Model Context Protocol)

A standard interface for extending Claude Code's toolbox by connecting it to external services, APIs, databases, or platforms.

 

MCP Server

A service that exposes tools to Claude Code via the MCP standard. Can be installed globally or per-workspace (mcp.json).

 

mcp.json

The workspace-level config file where MCP server configurations are stored when installed per-project.

 

Global vs Workspace Install

MCP servers can be installed globally (available in all projects) or scoped to a single workspace via mcp.json.

 

MCP Authentication

MCP servers may require auth tokens or OAuth flows.