Save Time Writing AI Prompts with Raycast Snippets

TL;DR

Use Text Replacements on iOS/macOS to map short shortcuts to longer phrases, reducing repetitive typing for templates and prompt blocks.

Briefing Cornell Notes

Briefing

Long, repetitive prompts are a major friction point when using AI chat tools like ChatGPT. The practical fix is to turn frequently used prompt fragments into reusable “snippets” (text expanders) so the full, well-structured instruction set gets inserted instantly—cutting typing time and improving answer quality by keeping prompts consistent.

On iOS and macOS, this capability is built in under Keyboard settings via Text Replacements. The workflow is straightforward: create a longer phrase and assign it a short shortcut. A common example is email entry. Instead of typing an email address repeatedly, the user assigns a short trigger (like “at5”) to a specific address, so the full email populates automatically. The same approach works for any recurring text—templates, short phrases, or longer prompt blocks.

For even more power, Raycast Snippets adds a dedicated snippet system on macOS. Raycast’s snippet library stores longer prompt templates that can be inserted anywhere using a hotkey (the transcript mentions calling Raycast with Option + S). One example is a “cooking assistant” snippet designed to make ChatGPT responses better by enforcing specific instructions. Rather than manually pasting a lengthy instruction prompt each time, the user types a short trigger (like “at Cook”), and Raycast inserts the entire structured prompt. The result is faster prompting and more reliable behavior because the model always receives the same instruction format.

Raycast also supports Dynamic text, which lets snippets incorporate additional context automatically. In the transcript, Dynamic text is used with Claude AI to generate newsletter email headlines. The user drafts content in Notion, copies it, then triggers a Raycast snippet (like “at headline”). Raycast inserts the prebuilt prompt and also pulls in the most recent clipboard content—shown as green text in the example—so Claude AI receives both the instruction template and the copied draft. The system then produces 10 headline options for brainstorming.

The key takeaway is that prompt reuse isn’t just about saving keystrokes; it also standardizes prompt structure, which tends to yield more consistent outputs. Whether using macOS/iOS Text Replacements or Raycast Snippets (especially with Dynamic text and clipboard integration), the workflow turns AI prompting into a quick, repeatable action rather than a fresh writing task every time.

Cornell Notes

Reusable text snippets can eliminate the need to type long AI prompts repeatedly. macOS and iOS provide Text Replacements under Keyboard settings, where a short shortcut expands into a longer phrase—useful for email addresses, templates, and prompt blocks. Raycast Snippets goes further by storing lengthy, structured prompt templates and inserting them via hotkeys. Its Dynamic text feature can also pull in context from the clipboard, letting a single snippet generate outputs (like 10 newsletter headlines) based on whatever draft was last copied. The payoff is both speed and consistency: the AI receives the same instruction structure every time.

How do Text Replacements on macOS/iOS reduce repetitive typing for AI and non-AI tasks?

Text Replacements lets users define a longer phrase and a short trigger. When the trigger is typed, the full phrase expands automatically. The transcript gives email entry as a concrete example: instead of typing an email address each time, a short code (e.g., “at5”) expands into the assigned address. The same mechanism can be used for reusable templates and prompt fragments.

What does Raycast Snippets add beyond built-in Text Replacements?

Raycast Snippets stores longer, ready-to-use text blocks—especially useful for lengthy AI prompts—and inserts them via a hotkey (Option + S is mentioned). The transcript’s “cooking assistant” snippet is an example: a long instruction set is inserted instantly when the user types a short trigger (like “at Cook”), so ChatGPT gets a consistent, structured prompt without copy-paste.

Why does using a structured snippet improve AI output quality in the transcript’s examples?

The cooking assistant snippet is designed to “follow these specific instructions.” By always inserting the same structured instruction text, the AI receives consistent guidance, which the transcript links to better results than manually crafting prompts each time.

How does Raycast Dynamic text work in the newsletter headline workflow?

Dynamic text inserts additional context alongside the snippet. In the example, the user copies a draft from Notion, then triggers a Raycast snippet (like “at headline”). Raycast inserts the prebuilt headline prompt and also includes the last copied clipboard content (shown as green), enabling Claude AI to generate 10 headline options based on that draft.

What role does the clipboard play in making AI prompting faster?

The clipboard integration removes an extra step: instead of pasting the draft into the prompt manually, Raycast automatically pulls in the most recently copied text. That means one action—triggering the snippet—can both supply the instruction template and attach the relevant content for the model to work on.

Review Questions

What are two different ways mentioned for creating reusable text expansions on Apple devices, and what does each one do well?
Describe how the cooking assistant snippet changes the prompting workflow compared with typing a full prompt every time.
In the newsletter headline example, what two inputs does Raycast combine when using Dynamic text?

Key Points

1
Use Text Replacements on iOS/macOS to map short shortcuts to longer phrases, reducing repetitive typing for templates and prompt blocks.
2
Raycast Snippets stores lengthy AI prompt templates so a short trigger can insert the full instruction set instantly.
3
Hotkey-driven snippet insertion (Option + S) makes prompting faster and more consistent than manual copy-paste.
4
Raycast Dynamic text can automatically include additional context, especially by pulling in the latest clipboard content.
5
Standardizing prompt structure (like the cooking assistant’s specific instructions) can lead to more reliable AI responses.
6
Clipboard-based workflows let users generate outputs (e.g., 10 newsletter headlines) from whatever draft was last copied with minimal extra steps.

Highlights

Text Replacements can turn repeated typing—like email addresses—into one keystroke expansion.

Raycast Snippets can inject a long, structured “cooking assistant” prompt with a short trigger, improving both speed and consistency.

Dynamic text plus clipboard integration lets one snippet generate newsletter headline options from the most recently copied draft.

The workflow shifts prompting from “write from scratch” to “reuse a template + add context,” cutting friction for every AI request.

Topics

Prompt Snippets
Text Replacements
Raycast Snippets
Dynamic Text
Clipboard Integration