Voice Library Operations
Voice Center
Browse cue coverage, review provider readiness, and keep editing separate from credential management and support tooling.
Main action
Refresh the queue, filter to the cues that need work, then generate or edit only the affected phrases.
Server:
Checking server...
TTS Key:
Checking key...
Preview / Approve:
Checking availability...
Browse and act on cues
Filters, search, and generation stay together as the primary operator lane.
Access and credentials
API keys and admin tokens stay separate from the browse flow so the main editing surface is less dense.
ELEVENLABS_API_KEY is missing. Audio generation will remain blocked until a key is configured.
Voice health
Loading...
App Live Audio Proof
Latest app ACK proof from the Live voice path.
Loading app audio proof...
Voice pack file board
One place to review every pack, manifest entry, MP3 file, missing cue, stale cue, and orphan file.
Loading voice file board...
| Pack | Status | cue_id | Kind | Text | File | Bytes | Updated | Action |
|---|
Quick guide
- Edit text, save it, and then generate audio only when the phrase is ready.
- Generate audio creates a new MP3 from the currently approved text.
- Play current previews the active audio already attached to the cue.
- Missing audio can be filled through Changed-only or Approve flows.
- Changed-only generates only cues that changed since the previous generation pass.
- Dry Run shows what would be generated before any files are written.
- Audio status can show current audio, preview-only audio, or missing audio.
- Manual updates highlight cues edited directly by an operator.
- Without an API key, generation remains unavailable even if editing still works.
Column widths
| Coach | cue_id | Meaning | Event | When | Approved | Edit | Words | Status | Audio | Actions |
|---|