Claude Command Reference

Projects, Artifacts, memory, styles, and connectors on claude.ai

2026 Edition

About this guide

This guide covers the core features of Claude.ai — the chat product available on web, desktop, and mobile. It does not cover Claude Code (the terminal-based coding agent) or Claude Cowork/Claude for Work (the agentic task-execution and team/enterprise product) — those have their own guides.

Sources for this guide are Anthropic's own Help Center (support.claude.com) and product pages (claude.com). Availability of any given feature depends on your plan (Free, Pro, Max, Team, Enterprise) and platform (web, desktop, mobile), and features roll out gradually — check your own account settings for the authoritative current state.

Starting a conversation
Projects
Artifacts
Memory
Custom styles
Connectors
Keyboard shortcuts
Plans and where features are available
Official resources

1. Starting a conversation

Type / inside an empty, focused message box to open the slash command and connector menu — from there you can pick projects, files, connectors, and other quick actions without leaving the composer.

The most-used in-conversation commands take effect immediately:

Command	What it does
`/clear`	Start a fresh conversation
`/undo`	Undo the last change in an editable artifact or document
`/compact`	Summarize a long conversation to continue with less context overhead

Because the exact command palette varies by plan, platform, and rollout, type / in your own composer to see what's available to you — that live menu is the source of truth, not a static list in a document.

2. Projects

Projects are self-contained workspaces within Claude, each with its own chat history and knowledge base — a way to keep related conversations, files, and instructions together instead of scattered across separate chats.

Core features:

Knowledge base: upload documents and files to give every conversation in the project shared context, plus custom project instructions to shape how Claude responds within that project.
Expanded capacity (Pro, Max, Team, Enterprise): paid plans get Retrieval-Augmented Generation (RAG), which can expand a project's usable knowledge capacity by up to 10x while maintaining response quality as it approaches context limits.
Free plan limit: free accounts can create up to five projects.

Collaboration (Team and Enterprise only):

Granular sharing with Can view and Can edit permission levels.
Share with individual members, add people in bulk by email, or share with the whole organization.
A Shared with me tab surfaces projects other people have shared with you.

3. Artifacts

Artifacts are a dedicated side window where Claude displays substantial, self-contained content, separate from the back-and-forth of the conversation. Claude creates an artifact when content is significant and self-contained — typically over 15 lines — and the kind of thing you're likely to want to edit, iterate on, or reuse, rather than just read once.

Common artifact types:

Documents (Markdown or plain text)
Code snippets
HTML websites
SVG images
Diagrams and flowcharts
Interactive React components

Working with artifacts:

Edit and iterate: ask for changes in the conversation, or, for Markdown documents, highlight specific text and choose Edit with Claude to make a targeted change without describing the whole document again.
Multiple artifacts per conversation: work with several at once and switch between them using the chat controls.
Artifacts sidebar: every artifact you've created is accessible from a dedicated Artifacts section, so you don't have to scroll back through the conversation to find one.
Export: view the underlying code, copy the content, or download the file.
Interactive/AI-powered artifacts: an artifact can embed Claude's intelligence directly — for example, a small app that answers questions or generates content for whoever's using it, not just the person who built it.

Prerequisite: artifacts require Code execution and file creation to be turned on in your account settings.

4. Memory

Memory lets Claude retain context across separate conversations, so it starts from what it already knows about you instead of every chat beginning from zero. Memory works the same way in regular chat and in Claude Cowork — wherever you work with Claude, it draws on the same stored context.

How it works:

Memory is on by default for Free, Pro, and Max plans, across web, desktop, and mobile.
Go to Settings → Capabilities → Memory to see everything Claude has stored, organized by topic, and to edit or delete any of it.
Start a new conversation and ask "What do you remember about me?" to check that memory is working and see what's stored.

Sensitive topics: by default, Claude does not store information related to health, race, ethnicity, religious beliefs, political views, gender identity, and similarly sensitive categories. Turning on include sensitive topics in memory lets Claude retain things like a food allergy so it can factor that into, say, recipe suggestions — but this is opt-in, not the default.

Import and export: Claude supports importing memory data — including from ChatGPT — and exporting your own Claude memories. Look for the import/export option under Settings → Memory.

5. Custom styles

Styles let you tailor how Claude responds — its tone, structure, and level of detail — without re-explaining your preferences in every conversation.

Preset styles:

Style	What it does
Formal	Clear, polished responses
Concise	Shorter, more direct responses
Explanatory	Educational responses aimed at helping you learn a concept

Custom styles: beyond the presets, you can create your own style by providing sample content that demonstrates the tone and format you want, or by uploading content that represents your desired voice. Open + Create & edit styles from the "Use style" menu in the composer to set one up.

Styles are useful for keeping output consistent across different needs — e.g. technical explanations for developers, on-brand copy for marketers, or a consistent voice across a team.

6. Connectors

Connectors link Claude to external tools and data sources so it can pull in real context — email, documents, project trackers — instead of relying only on what you paste into the chat. Available connectors and what they can access vary by plan and by what your organization has enabled.

Common connector categories include productivity and file storage (Google Drive, Google Calendar), communication (Gmail, Slack), and business tools (HubSpot, Asana, and others) — see the Connectors section of your Claude.ai settings for what's available and how to connect an account.

7. Keyboard shortcuts

Shortcut (Windows/Linux)	Shortcut (Mac)	Action
`Ctrl+Shift+O`	`Cmd+Shift+O`	Start a new chat
`Shift+Tab`	`Shift+Tab`	Toggle plan mode
`Ctrl+C`	`Cmd+C`	Cancel a running response
`Ctrl+R`	`Cmd+R`	Search session history

Type / in an empty composer for the full, current command menu — including any additional shortcuts available for your plan, platform, and rollout stage.

8. Plans and where features are available

Feature availability is plan- and platform-dependent, and Anthropic ships changes on a rolling basis. As a general shape:

Free: chat, a limited number of projects (up to five), memory (with the sensitive-topics default off).
Pro / Max: expanded project capacity via RAG, full memory controls, custom styles.
Team / Enterprise: everything above, plus project sharing with granular permissions, organization-wide connector controls, and admin-managed settings.

Always confirm current availability in your own account — the menus and settings pages are the authoritative source, not any static document (including this one).

9. Official resources

Topic	URL
Getting started with Claude.ai	https://support.claude.com/en/articles/12997377-getting-started-with-claude-ai
What are projects?	https://support.claude.com/en/articles/9517075-what-are-projects
How can I create and manage projects?	https://support.claude.com/en/articles/9519177-how-can-i-create-and-manage-projects
What are artifacts and how do I use them?	https://support.claude.com/en/articles/9487310-what-are-artifacts-and-how-do-i-use-them
Publish and share artifacts	https://support.claude.com/en/articles/9547008-publish-and-share-artifacts
Chat search and memory	https://support.claude.com/en/articles/11817273-use-claude-s-chat-search-and-memory-to-build-on-previous-context
Connectors	https://support.claude.com/en/collections/15399129-connectors
Anthropic Help Center	https://support.claude.com/en

Final accuracy note

This guide reflects Anthropic's own Help Center documentation as reviewed on September 3, 2026. Claude.ai ships changes frequently and the exact command menu, feature set, and plan boundaries vary by account, platform, and rollout. When this document and the live product differ, the product itself — and the official Help Center at support.claude.com — is the source of truth.

Claude Command Reference — 2026 Edition