The right mental model before you start
Claude is Anthropic's AI model, built by former OpenAI researchers who took a different approach: safety first, capability second. That's not just marketing — it directly shapes how Claude reasons, follows instructions, and behaves under pressure.
Most people use Claude across three broad categories:
Claude Sonnet 4 is the optimal speed/capability balance for most tasks. Claude Opus 4 when you need deep reasoning. Claude Haiku 4 for simple, high-volume tasks.
Claude is not a search engine or autocomplete. It's a collaborator with reasoning capability. Treat it like a smart colleague you brief well: the more context you give it, the better it works.
The most common mistake: one-line prompts expecting magic. The reality: the best results come from prompts that explain the context, objective, constraints, and expected format.
In this course you'll learn to use Claude Code — Anthropic's official CLI that integrates Claude directly into your terminal and development workflow.
ejercicio
Your first full-context prompt
Open claude.ai or your terminal with Claude Code. Copy this prompt and observe the difference vs a generic one-liner.
Act as an expert AI productivity consultant. Context: I am a [your role] and I primarily work in [your area]. I spend a lot of time on [repetitive task you want to optimize]. Task: Explain in 3 concrete steps how I could use Claude to reduce the time I spend on this task. Include a sample prompt for each step. Format: Numbered list, maximum 150 words per step, no unnecessary jargon.