Getting Around the App — Interface and Keyboard Shortcuts

Series: Claude Learning Journey · Getting Started

The Claude Mac desktop app is intentionally straightforward. A conversation window, a text input, and a send button. That simplicity is the point — you should not need to learn a complex UI to have a useful conversation. But there are features and shortcuts that make the experience faster and more powerful once you know they exist.

This post covers the desktop app interface and the keyboard shortcuts that save time. It is written for desktop app users first. If you are using the CLI, the equivalent commands are noted alongside each section.

The main conversation window

The conversation window shows your current thread with Claude. You can scroll back through previous exchanges, click on any previous message to continue from that point, and start new conversations from the sidebar.

The sidebar shows your conversation history grouped by date. You can search across all conversations with the search field at the top. This is useful when you remember having a useful exchange but cannot recall which conversation it was in.

Starting new conversations

Click the new conversation button (top left) or use a keyboard shortcut. A new conversation starts fresh — no memory of previous threads, no context carried over.

You can also right-click on a previous conversation to:

• Rename it (useful for tracking ongoing projects)
• Delete it
• Duplicate the starting prompt into a new conversation

Switching between models

The model selector is in the top right of the conversation window. It shows your current model (Sonnet, Opus, Haiku, etc). Click it to switch. The switch takes effect immediately and applies to your next message.

This is useful when you want Opus-level reasoning for a complex problem but Sonnet for quick, everyday questions. You do not need to start a new conversation to switch models.

If you are using the CLI, switch models with:

/model opus

Or specify the model when you start a session:

claude --model opus

Sending messages

Type in the input field and press Enter to send. Shift-Enter adds a line break if you want to write a multi-line message before sending.

You can also paste images directly into the input field. Claude will attach them to the message and reason about them. This works with screenshots, photos, and documents.

Claude Code users: the equivalent is attaching file paths to your prompt or using the --attach flag.

Editing your message before sending

Hover over your sent message and click the small edit icon. This lets you revise what you sent without the context of a new message being added to the conversation. Useful when you realise you phrased something poorly after hitting send.

Regenerating a response

If Claude’s response is not useful, click the regenerate button (the circular arrow icon) next to Claude’s reply. Claude will generate a new response to your last message. This is useful when you want a different angle or when Claude misread your intent.

Keyboard shortcuts that save time

Cmd+Enter — send message (instead of hitting the send button) Cmd+Shift+C — copy last Claude response to clipboard Cmd+K — clear current conversation (same as /clear in CLI) Cmd+N — new conversation Cmd+F — search within current conversation Escape — cancel a response that is still generating

These take a few minutes to become automatic. After that, you will not want to use Claude without them.

Slash commands in the input field

The desktop app supports text shortcuts in the input field:

• Type /new to start a new conversation
• Type /clear to clear the current conversation
• Type /model and select from the model picker

These are the same commands available in the CLI, adapted for the desktop app interface. They are useful muscle memory once you start using Claude regularly.

Attachment tips

You can attach files by dragging them into the conversation window, or by clicking the paperclip icon in the input field. Claude handles most common formats: PDF, images, code files, plain text.

For code files specifically, attaching the file directly gives Claude better context than copying and pasting the contents. The file retains its structure and formatting.

If you are on the CLI, attach files with:

claude "Explain this" --attach path/to/file.py

What not to do

Do not paste API keys or secrets into the conversation. The conversation may be logged depending on your account settings. Use environment variables or proper secrets management, not chat.

Do not assume Claude knows your project without providing context. It has no access to your files unless you attach them or share relevant code.

Do not ask vague questions expecting specific answers. “Should I use a database here?” is not useful. “What are the trade-offs of PostgreSQL vs DynamoDB for a write-heavy event log with strong consistency requirements?” gives you something concrete to work with.

Try it yourself

Open the Claude app and try the following in sequence:

1. Start a new conversation
2. Attach a file from a project you are working on
3. Ask a specific question about it
4. Switch to a different model and ask the same question
5. Notice the difference in the response

This gives you a feel for the interface without any commitment. The investment is about 2 minutes.

What’s Next

With the basics covered, Stage 2 dives into the workflows that make you productive. Next up: prompt engineering — the skill that multiplies everything else you do with Claude.

Part of the Claude Learning Journey series · Next: Prompt Engineering That Actually Works