You are reading Nightly documentation for 0.12.4.dev0+g50bde75.

This documentation may describe behavior that differs from Stable.

Open Stable documentation

Documentation version

0.12.4.dev0+g50bde75 · Nightly
Stable Beta Nightly

Sounds

Sounds

Holaryn Agent can play a short sound when something happens — an answer arriving, a permission request, a question, dictation starting. Everything is off by default; Settings → Sounds is where you turn events on and choose what they play.

How it works

Sounds play in your browser (or the desktop app's window). The page has a master Play sound effects switch and a Volume slider with a Test button; below them, each event has its own checkbox and sound picker, grouped into Activity, Needs your attention, and Dictation. Enabling an event just picks it a sound — changes save immediately and take effect without a reload.

The events

Event Group Default sound
Answer ready — the model finished and returned an answer Activity Chime
Thinking (progress tone) — repeats on an interval while the model is working Activity Tick
Approval needed — the agent is waiting for a permission decision Needs your attention Bell
Question asked — the agent asked you a question Needs your attention Ding
Error — the agent hit an error Needs your attention Soft buzz
Plan ready for review — a staged plan is awaiting your approval Needs your attention Chord
Code plan proposed — a code session proposed a plan for review Needs your attention Chord
Security alert — a prompt-injection attempt was flagged Needs your attention Alarm
Dictation started listening — dictation began capturing your voice Dictation Rising
Dictation finished — dictation stopped and is transcribing Dictation Falling

The thinking tone has its own Thinking interval control: how many seconds between repeats, from 3 to 60 (default 8).

The built-in sounds

Every event's picker offers three groups:

  • Tones — synthesized in the browser, no files: Chime, Blip, Rising, Falling, Soft buzz, Tick.
  • Pack — a bundled WAV pack served with the UI: Bell, Ding, Pop, Chord, Alarm.
  • Custom — your own uploaded clips.

Audition while you browse

The picker plays each sound as you hover over it or arrow through it with the keyboard, so you can hear your options before committing — nothing is assigned until you actually pick one. Each event row also has a Preview button that plays its currently assigned sound on demand.

Custom sounds

Under Custom sounds, choose Browse for a sound… to upload your own clip:

  • Formats: WAV, MP3, or OGG.
  • Size limit: 1024 KiB per clip (these are short effects, not music).

Uploaded clips are stored in the host state directory and appear in every event's picker under Custom, with Preview and Remove buttons. Removing a clip returns any event that used it to that event's default sound.

  • Web Interface — the UI these cues play in.
  • Voice — dictation, whose start/stop cues live here.
  • Settings — the Settings page reference.
← Previous Slash commands Next → Subagents and coding jobs