The Action Extractor
Ramble about the project, walk away with a checklist.
Pull the action items and decisions out of what I said. Format them as a checklist - one per line, verb first - followed by a short list of open questions. Nothing else.
Use case
Bind your own prompts to shortcuts so routine writing tasks take one hotkey.
Starter spell: The Action Extractor · suggested ⌃⌥6
The moment
Some writing tasks are the same shape every time - standups, follow-ups, weekly summaries. The real win is not just dictation; it is binding the shape itself to a shortcut, so the same prompt runs on whatever you say.
The Action Extractor is the starter template that pulls the checklist out of a spoken update - copy it from the library or write your own in the app.
One recording, both versions
Raw transcript
ok so today, uh, the API migration is mostly done but staging is still broken, Lena's on the certificates. we decided to ship the beta on the twentieth. um, I need to email the two pilot customers, and nobody knows who owns the docs
Whisper heard every “um”.
The Action Extractor · Example
☐ Finish the API migration ☐ Fix the staging certificates (Lena) ☐ Email the two pilot customers ☐ Ship the beta on the 20th (decided)
Open questions • Who owns the docs?
Pasted at your cursor · both versions kept in History
How it plays out
Any Mac app with a text field: mail, chat, a doc, a form. WhisperWizard remembers which app was in front.
One hotkey starts the recording and picks The Action Extractor. Speak the way you would tell a colleague.
Whisper transcribes on your OpenAI account, then the spell's prompt runs on the transcript.
The text is pasted where your cursor was, your previous clipboard is restored, and History keeps both versions with the audio.
The starter spell
Ramble about the project, walk away with a checklist.
Pull the action items and decisions out of what I said. Format them as a checklist - one per line, verb first - followed by a short list of open questions. Nothing else.
Tips
Write the format into the prompt (checklist, standup, summary) and every recording comes back in it.
Each template picks its own OpenAI model, so quick tasks and careful ones can run on different ones.
There is no limit on templates. Give each recurring task its own hotkey.
Questions
As many as you want. Every plan includes unlimited custom templates, each with its own prompt, model and shortcut.
Yes. ⌥Space opens the template palette: type to search, use the arrow keys, press Enter to pick one, then speak.
Five more

Dictate the reply you would have typed and get polished text back, ready to send.

Get a first draft out of your head and onto the page, then refine it with a custom template.

Speak in one language and let a template translate or reshape your words for you.

Capture thoughts the moment they happen - every recording is kept for instant replay.

Rest your hands and wrists while your voice keeps the words flowing.
WhisperWizard is macOS only (10.13 or newer on Intel, 11 or newer on Apple Silicon) and runs on your own OpenAI API key - your recordings go straight to OpenAI, we never see or store them on our servers (History keeps its copy on your Mac). The honest comparison says when built-in dictation is already enough.
Every plan includes unlimited custom templates, recording history, and all features - one-time purchase, lifetime usage, and a 14-day money‑back guarantee.
From $29 once · your own OpenAI key · 14-day refund