Getting Started with Claude Agents
A hands-on introduction to Claude agents.
$ open "Welcome to the lab"
Run your first agent task and learn the three-pane lab.
checkpoints run as you work
✓ checkpoint passed, lab 2 unlocked
What you'll build
Curriculum
10 labs, about 1.5 h
- 1
Welcome to the lab
Run your first agent task and learn the three-pane lab.
5 min
Pro - 2
Watch the agent work
See how the agent reads, acts, checks, and continues.
7 min
Pro - 3
Give clear instructions
Write prompts with a goal, context, constraints, and deliverable.
8 min
Pro - 4
Plan before you build
Have the agent inspect first and propose a small plan.
7 min
Pro - 5
Continue the conversation
Use short follow-up prompts that build on earlier context.
9 min
Pro - 6
Add project instructions
Use CLAUDE.md to give the agent reusable project guidance.
10 min
Pro - 7
Use a custom tool
Decode a secret message with real MCP tools from a plugin.
9 min
Pro - 8
Use a skill and command
Inspect, then invoke a plugin skill and a custom slash command.
10 min
Pro - 9
Delegate to a helper
Inspect a read-only subagent, then delegate a review to it.
9 min
Pro - 10
Your graduation mission
Combine tools, a skill, and a helper to finish one real mission.
12 min
Pro
Your instructor
Elvis Saravia
Founder, DAIR.AI
Elvis founded DAIR.AI and wrote the Prompt Engineering Guide, one of the most widely used references on working with language models. He has worked at companies like Meta AI and Elastic, and specializes in AI agents, RAG, and harness engineering.
How labs work
Live workspace
You run the real tools in a browser workspace. Nothing to install.
Checkpoints
Each lab checks your files and output, then unlocks the next one.
Your work stays
The workspace persists between labs so you finish with a project.