ClaudeHQ Foundations

Meet Claude

By the end: you'll know what Claude is, how it's different, and which gear to reach for.

Claude is an AI assistant made by Anthropic. You tell it what you need in plain English, and it writes, explains, analyzes, summarizes, brainstorms, or builds things alongside you.

Most people treat AI like a search box — type a few words, hope for the best. Claude works better when you treat it like a sharp assistant you're handing a task to. Small shift, big difference. The whole course is built on it.

Claude comes in a few versions, and you don't need to memorize their names. Think of them as three gears:

• The fast model — quick drafts, summaries, simple rewrites. Speed over depth.
• The balanced model — your everyday default. Most writing, analysis, and work lives here.
• The strong model — the heavy lifter. Reach for it when the task is complex, high-stakes, or has a lot of moving parts.

Start with the balanced one. Move up when an answer feels thin, move down when you just want speed. (Want exact names and specs? The models table on ClaudeHQ has them — but you can use Claude for months without ever thinking about it.)

Your First Real Prompt

By the end: you'll turn a vague request into one that actually works — every time.

A prompt is just the instruction you give Claude. This module teaches the one formula that fixes most weak results.

When Claude hands you something generic, it's usually not being dumb — it's guessing, because you didn't give it enough to work with. Better instructions, better output. That's almost the whole game early on.

Every strong prompt has three parts:

• What — the job. ("Write a LinkedIn post announcing our training.")
• How — the style or role. ("Make it practical and beginner-friendly, no hype.")
• Constraints — the limits. ("Under 180 words, strong first line, soft call to action at the end.")

Think of it like ordering at a deli. "Sandwich" gets you whatever they feel like making. "Turkey on rye, toasted, no mayo, extra pickles" gets you lunch. Claude's the same — tell it what, how, and what to leave off.

Weak: "Write a LinkedIn post."
Better: "Write a LinkedIn post announcing our free Claude training. Practical and beginner-friendly, no hype. Under 180 words. Open with a hook, close with a soft call to action."

Same tool. Completely different result.

Make It Actually Useful

By the end: you'll know the four dials that turn a decent answer into one you'd actually send.

Module 2 got you a good answer. This one gets you the right answer — shaped to sound like you, land with your audience, and arrive in the format you need.

A generic answer is easy. A useful one is shaped before Claude writes a word. Four dials do most of the shaping.

• Persona — who Claude should act like. ("Act like an experienced operations manager.")
• Tone — how it should sound. ("Calm, direct, no corporate fluff.")
• Format — what shape the answer takes. ("Subject line, one short paragraph, three next steps.")
• Examples — what "good" looks like. (Paste one in and say "match this style.")

If Module 2 was ordering at the deli, this is deciding whether a line cook, a chef, or a caterer makes it. Same ingredients, very different result.

Instead of: "Rewrite this email."
Try: "Rewrite this email as an experienced operations manager. Keep it calm, clear, and professional. Format it as a subject line, one short paragraph, and three bullet next steps. Plain English, no fluff."

Don't Get Burned

By the end: you'll know exactly where to trust Claude and where to check its work.

Claude is fast and capable. It's also not automatically right. This module is about knowing the difference.

Claude can be confident and wrong at the same time. It can sound polished while making something up. If you're using it for anything that matters, your job isn't just to prompt well — it's to verify smart.

Treat Claude like a sharp junior teammate. Great for drafts, summaries, brainstorms, reshaping things, and analyzing what you give it. Not your final word on facts, law, money, or medicine.

Lean on Claude to: draft, summarize, brainstorm, reorganize, transform, and analyze material you hand it.

Don't blindly trust it to: invent citations, give final legal or medical advice, replace an expert, or be right just because it sounds sure.

After Claude answers something that matters, ask:

• "What are you least sure about here?"
• "What did you assume?"
• "How would I check this?"

Your Workflow Starts Now

By the end: you'll have used the three features that turn Claude from a chat box into part of how you actually work.

Up to now you've been typing into a blank chat. This module shows you the features most people never find: Projects, file uploads, and artifacts.

A blank chat is fine for one-off questions. Real work is repeat work — same context, same files, same kind of output. These three features are how you stop starting from scratch every time.

• Projects — a home for one body of work. Keep your files and instructions in one place, so every chat inside it already knows the context. No more re-explaining yourself.
• File uploads — give Claude the actual material. Drop in a PDF, a spreadsheet, a doc, your notes — and it works from your real stuff instead of guessing.
• Artifacts — when Claude makes something you'll reuse (a document, a summary, a simple web page), it builds it in its own window you can edit, copy, and download.

The shift: beginners ask Claude to "help" in the abstract. Real users hand it the actual brief, the actual file, and ask for a specific thing back.

Go deeper on Projects in the full guide →