Voice and memory
Listen, talk, and inspect what a folk remembers.
Spoken replies
Add an ElevenLabs key in Settings, choose a voice, and use the speaker control on a response. Auto-speak can read new replies as they arrive.
The harness rewrites code-heavy Markdown into speech-friendly text before synthesis. Keys stay on the harness and never reach the renderer.
Call mode
On macOS, call mode combines on-device speech recognition with the configured ElevenLabs voice. It is half-duplex: the microphone pauses while the folk speaks so the app does not transcribe itself.
Tool progress and approval requests can be narrated during a call so long work does not go silent.
Dictation
The composer microphone uses Apple's local speech recognition in the macOS desktop build. Linux and Windows dictation are not shipped.
Memory
Each folk can keep plain-text memory you can inspect, correct, or delete. It is not a hidden remote profile. It lives with the local OpenFolks data.