> ## Documentation Index
> Fetch the complete documentation index at: https://www.plain.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# AI slash commands

> Get writing help from Plain AI in the composer with a slash command.

Use AI slash commands to get writing help from Plain AI directly in the composer, no context switching needed.

<Frame>
  <img src="https://mintcdn.com/plain/esnLSCL6AbjPxXFS/public/images/ai-slash-commands.png?fit=max&auto=format&n=esnLSCL6AbjPxXFS&q=85&s=8ace1ff4627f87ab345e523d8d69c06b" alt="AI slash commands in Plain" width="2280" height="1200" data-path="public/images/ai-slash-commands.png" />
</Frame>

## How it works

While drafting a reply in a support thread, start a new line with <kbd>/</kbd> followed by your instruction, then press **Enter**. Plain AI will reference the last 5 messages in the thread alongside whatever you have already written in the composer, and replace the composer content with a ready-to-edit draft.

Slash commands work for:

* Drafting a reply from scratch

* Editing or rewriting something you have already written

* Adjusting tone, length, or clarity

## Examples

These are examples. You can give it any prompt:

* `/fix typos`

* `/ask for a screenshot and workspace ID so we can debug`

* `/ask what version of the SDK they are using`

* `/rephrase this to sound more helpful`

* `/summarize and confirm the issue before escalating`

## Things to know

* The slash command must be on its own line at the top level of the composer. It will not trigger inside a list or block element.

* The AI response replaces the entire composer content, including any text you had written before the slash command.

* Only the last 5 messages in the thread are included as context. Earlier messages in a long conversation will not be referenced.

* If the request fails, you will see a "Failed to generate or edit message" notification. Retry by typing the command again.

* Requires Plain AI to be enabled in your workspace settings.
