AI Guides

Claude Features Cheat Sheet for Power Users

Mika Reyes
Mika Reyes

Co-founder at King’s Cross Labs · ex-LinkedIn PM & Forbes 30 Under 30

Most people treat Claude like a search bar. Power users build a system.

These seven features stack: one saved command at the bottom, a job that runs without you at the top. Bookmark this. Claude keeps shipping features and this map still holds.

If you only chat, you're leaving most of the product on the table. Commands, artifacts, skills, projects, connectors, plugins, and scheduled routines give the same model a playbook, access to your files, and a calendar. You don't need all seven today. You need to know what each one is for so you stop rebuilding the same setup in every chat.

The foundation: stop retyping, start shipping

1. Commands (slash commands)

Slash commands are prompts you save and trigger with one word, like /summarize. Stop rewriting the same meeting-recap or email-draft instructions. Store them once, call them when you need them. If you paste the same setup paragraph more than twice a week, make it a command. For Claude Code session cleanup patterns, I also cover a few favorites in 3 Claude Commands to Manage Your Sessions.

2. Artifacts

An artifact is the deliverable, not a wall of chat. Claude builds a doc, deck, chart, or working app you can edit live in the side panel. The output looks like finished work, not notes you'll clean up later. Ask for an artifact whenever the result needs to leave the conversation. Full walkthrough: Intro to Claude Live Artifacts.

Key insight: Chat is for thinking. Artifacts are for shipping.

The playbook layer: teach Claude once

3. Skills

A skill is a packaged how-to: steps, examples, your style. Claude follows it automatically when a task matches, even if you never type the skill name. You're handing Claude your SOP once instead of re-explaining it every Monday. Use skills for work you do the same way every time: brand voice, research format, weekly reporting. Still fuzzy on the concept? Read What is a Skill?. Ready to build one? Use How to Create Your Own Custom Skill.

4. Projects

A project keeps your files, instructions, and context in one place. Every chat inside it already knows the business. No more re-uploading the same brand deck or client brief. Use one when you keep pulling from the same files: a client account, a product launch, your personal ops hub. This is also where giving Claude memory starts to feel real, because the context lives with the work.

The system layer: connect, bundle, automate

5. Connectors

Connectors plug Claude into Google Drive, Slack, Notion, and similar tools so it reads live data instead of guessing from what you pasted. Use them when the answer already lives in your files or messages. If Claude keeps asking you to paste a Drive doc you already have, connect Drive. For creative-tool examples of the same idea, see 7 New Creative Tools as Claude Connectors.

6. Plugins

Plugins bundle commands, skills, and connectors into one install. You get a pre-built setup for coding, research, or ops without assembling each piece yourself. Reach for a plugin when you're starting a new job-to-be-done and want the whole kit, not one prompt.

7. Scheduled routines

Set the job once. Claude runs it on a schedule: daily reports, lead enrichment, inbox triage. No button-pressing from you. Use this when a task repeats on a predictable cadence and you're tired of remembering to start it. Same work you used to do in chat, now with a calendar invite.

How do you stack these without getting overwhelmed?

Start with one command and one artifact. Turn the prompt you keep reusing into a skill. Drop related files into a project. Connect the tools that hold your real data. Then look at plugins and scheduled routines.

People who try to make Claude do everything on day one usually bounce. The boring sequence wins: save the prompt, ship the deliverable, teach the playbook, then automate the repeat.

Still setting Claude up from scratch? Pair this cheat sheet with How to Setup Claude or How to Setup Claude for Small Business.

Here are some related guides to check out:

  1. What is a Skill?
  2. How to Create Your Own Custom Skill
  3. Intro to Claude Live Artifacts
  4. How to Give Claude Memory (Beginner to Advanced)
  5. How to Setup Claude (10 Minute Setup Guide)

Frequently asked questions

Can I use these Claude features without being technical?
Yes. Commands, artifacts, skills, projects, and connectors are built for normal Claude users, not engineers. You get most of the value by saving prompts, putting files in a project, and connecting Drive or Slack. Plugins and scheduled routines come later, once the basics feel natural.
What's the difference between a Claude skill and a slash command?
A slash command is a shortcut you choose to run. A skill is a playbook Claude can follow automatically when the task matches, even if you never type the skill name. Commands are great for intentional repeats. Skills are better when you want Claude to apply your process without being asked every time.
Do I need projects if I already use skills?
Yes, if you keep working from the same files and context. Skills teach Claude how to do the work. Projects give Claude the documents and instructions for that specific client, product, or ongoing body of work. They solve different problems and stack well together.
When should I use a plugin instead of building skills myself?
Use a plugin when you want a full pre-built setup for a new job without assembling commands, skills, and connectors one by one. Build your own pieces when the workflow is specific to your business and off-the-shelf bundles will not match your process.