How to save custom AI prompts (easy method)

Custom AI prompts can be saved and reused instantly inside Apple Notes-style workflows, eliminating the constant copy/paste and app-switching that slows down repeat tasks. The core payoff is speed: once a prompt is stored, it can be run directly on highlighted text (including voice notes), producing outputs like formatted documents, task lists, or research-backed drafts without rewriting the same instructions every time.

The workflow centers on cloning an existing prompt and giving it a new name, rather than overwriting the original. After selecting a prompt template (for example, a “formatting” prompt), the user copies the prompt text into a new prompt entry. From there, the prompt is saved under a chosen label—such as “format exercises”—so it can be triggered later on new notes. This approach matters because many useful prompts are long and detailed (like a project proposal formatter), making manual repetition impractical and error-prone.

Several real-world prompt examples illustrate what gets saved. One custom prompt was used to generate project proposals for an outbound email agency, with instructions that enforce specific formatting. Another prompt is designed for research: when drafting a blog post or social media content, it finds supporting evidence and returns citations to include. There’s also a prompt for turning video and podcast notes into usable text outputs. The transcript emphasizes that these prompts become most valuable when they’re tied to the notes the user already collects—voice notes, highlighted snippets, or rough drafts—so the AI can act on the user’s existing material.

To run a saved prompt, the user highlights the text inside the notes and invokes the assistant (the transcript mentions using Command J). The prompt then processes the selected content. If the notes are empty or lack context, the user may insert temporary filler text so the prompt has something to operate on; once the prompt is tested, the filler is replaced with the real content. The method is framed as “like saving and running custom AI prompts directly within Apple Notes,” turning notes into a control panel for repeatable AI tasks.

Finally, the transcript notes that a spreadsheet of saved prompts is available for sharing, and it points viewers to Reflect (the app referenced) plus additional note-taking prompts and templates. The overall message is straightforward: clone, name, save, and run—so custom AI instructions become a reusable tool rather than a chore.