A Quiet Dictation Keypad for Zed, Not a Silent One

WhisperKey costs $87.90. It is a clear handheld voice keypad intended to sit near the mouth, with four controls marked microphone, tick, cross and Enter. The current product facts do not confirm a microphone, key outputs, audio connection, remapping or Zed compatibility. It cannot make spoken prompts silent.

How do you dictate into Zed's Agent Panel today?

Zed's current official Agent Panel documentation explains typing into its message editor but does not document an Agent Panel dictation feature. An operating-system or approved third-party dictation service may insert text where the cursor is active. Availability, privacy and behavior depend on that separate service.

Prove the software path with the existing microphone before adding a handheld controller. Use invented content and follow workplace rules for audio processing.

What makes close-talking quieter than a desk mic?

A microphone nearer the mouth may allow softer speech at a suitable gain, but it does not prevent coworkers from hearing or other devices from recording. Room noise, microphone pickup and speech recognition vary. No acoustic performance is documented for this product.

Ask someone at the intended distance whether a test phrase is intelligible. For secrets, type or move to a private room.

What do the tick and cross keys do while reviewing?

Their printed symbols do not prove output. They should not map to Zed's keep, reject, Allow Once, Deny Once or Enter actions. Those decisions require the current diff, requested tool and permission scope on screen.

A verified microphone key might control push-to-talk in the separate dictation application. Leave the remaining positions inactive until a frequent reversible use exists.

Where does dictating code-heavy prompts still fail?

Recognition often struggles with punctuation, capitalization, file paths and similar-sounding identifiers. A grammatically neat transcript can name the wrong symbol or produce a valid dangerous command. Always inspect the full message before sending it.

Compare typing and dictation on harmless samples containing mixed-case names, dotted files and spoken punctuation. Count corrections, capture failures and privacy constraints. Confirm the recording indicator comes from the operating system or dictation software, not a printed icon. If release does not stop capture, disable the setup and use the normal control.

The WhisperKey is only a candidate after the voice software works and the hardware interface is documented. “Close to the mouth” is a position, not a guarantee of silence, privacy or accurate code. Run the same sample after sleep, microphone reconnect and application restart. Confirm the chosen audio input each time, because an unavailable device may cause software to fall back to another microphone. Delete recordings or transcripts under the relevant retention policy. A keypad does not change where a dictation provider processes audio, so review that boundary separately from the physical purchase.