Documentation Index
Fetch the complete documentation index at: /llms.txt
Use this file to discover all available pages before exploring further.

Skip to main content

Anthropic Training home page

Claude Enterprise
Claude Desktop

Getting Started

Welcome
How to use
Goals

Foundations

Overview
AI at work
Mental model
Fluency loop
When not to
Lab: Diagnose

Claude Landscape

Overview
Web vs Desktop
Surfaces map
Add-ins
Sharing
Lab: Three ways

Models & Modes

Overview
Choosing models
Thinking
Tool use
Agentic loop
Stack vocabulary
Lab: Compare

Context & Cost

Overview
Context is prompt
Where context lives
Hygiene
Token literacy
Compression
Lab: Project brief

Real Work

Overview
Deep research
Enterprise search
Analysis
Docs & decks
Ops & admin
Comms
Lab A: Research
Lab B: Live surface

Projects & Artifacts

Overview
Project setup
Artifacts
Handoffs
Lab: Durable Project

Skills

Overview
What is a skill
Using skills
Caveman skills
Building skills
Skills vs sub-agents
Lab: Ship a skill

Connectors & Cowork

Overview
Connectors
Cowork
Failure modes
Lab: Least privilege

MCP

Overview
What is MCP
MCP vs APIs
Tools & resources
Using vs building
Choosing options
MCP safety
Lab: Configure MCP
Stretch: Custom MCP

Agents

Overview
Loops in practice
Managed agents
First managed agent
Sub-agents
Code & tools
Lab: Managed agent
Stretch: Code tool

Evaluation

Overview
Judging outputs
Iterate
Different bars
Lab: Rubric

Safety & Ethics

Overview
What to share
Enterprise sharing
Ethics
Lab: Red-team

Capstone

Overview
Analyst pack
Ops pack
Power pack

Resources

Overview
Cheatsheets
Glossary
Further learning

Models & Modes

What tool use is

Tools let Claude act — browse, retrieve, compute, update — not only write.

What thinking is and is not

The agentic loop: plan, act, observe, revise

⌘I

Powered byThis documentation is built and hosted on Mintlify, a developer documentation platform