Skip to content

Author Slash Commands

Slash commands let you define reusable prompt templates that you can invoke from the composer with a /name prefix. Instead of retyping a long prompt you use regularly, you write it once, give it a name, and run it from the composer in any conversation.

A custom slash command has three parts:

  • Name: the token you type after the slash, for example review. Must start with a letter and can contain letters, numbers, hyphens, and underscores. Maximum 32 characters.
  • Description: a short label shown in the slash menu when you type /.
  • Template: the prompt text that gets inserted into the composer when you run the command. Can include {placeholder} or {{placeholder}} patterns for parts you want to fill in at run time.

Commands are stored in the app’s config layer (slash.customCommands) and persist across restarts.

  1. Open Settings > Slash Commands.
  2. Click New command.
  3. Fill in the name, description, and template.
  4. Save.

The command is immediately available in every conversation’s composer.

Type / in the composer to open the slash menu. Your custom commands appear alongside the built-in ones, labeled with their descriptions. Click one or finish typing the name and press Enter or Tab to expand it.

The template text replaces the /name in the composer. If the template has {placeholder} tokens, they remain in the composer text for you to fill in before sending. If the template has no placeholders, the full text is inserted ready to send.

Placeholders follow the {name} or {{name}} syntax (both work):

Summarize this in {language}: {{content}}

When you run the command, the template is inserted with the placeholders intact so you can type the values inline before sending. At present, placeholders are positional fillers; they are not prompted automatically.

Go to Settings > Slash Commands. Each command in the list has an edit (pencil) icon and a delete (trash) icon. Editing reopens the command editor; deleting asks for confirmation.

A small set of names is reserved for built-in commands and cannot be used for custom ones: btw, open, copy, export. If you try to save a command with one of these names, the editor will show a validation error.

How these differ from Core CLI slash commands

Section titled “How these differ from Core CLI slash commands”

The desktop slash commands described on this page are user-authored prompt templates that live in the desktop app. They are separate from the slash commands available in the Wayland Core CLI (wayland-core), which are skill-bound and execute directly inside the CLI’s agent loop.

If you are working with the Core CLI, see Slash Commands (Core).