Narration Script Template

Pipeline Steps

1. Prepare Material

Same as other templates — material from dispatcher.

2. Generate Script

Write a spoken-word script following style.md:

Hook opening
2-4 talking points with clear transitions
Strong closing

Apply any user style directives from .listenhub/creator/styles/narration.md (if exists) and sessionStyle (from style reference) on top of the baseline style. sessionStyle takes priority over the user style file, which takes priority over style.md.

Save as script.md in the output folder.

3. TTS Audio (Optional)

TTS is offered only if:

The pipeline requires it (user asked for audio)
LISTENHUB_API_KEY is available

If generating audio:

Select speaker:
- Check preferences.narration.defaultSpeaker in config — if set, use it
- Otherwise use built-in defaults from shared/speaker-selection.md:
  - Chinese: "原野" (CN-Man-Beijing-V2)
  - English: "Mars" (cozy-man-english)
- On first TTS use, ask the user via AskUserQuestion if they want to choose a different speaker. Save their choice to preferences.narration.defaultSpeaker for future runs.
Call TTS API:

listenhub tts create --text "$(cat /tmp/lh-content.txt)" --speaker "$SPEAKER_ID" --json \
  | jq -r '.data' | base64 -D > "{output}/audio.mp3"

Note: TTS max input is ~10,000 characters. For longer scripts, this is still well within limits for narration (typically 300-2000 chars).

If TTS fails: deliver script without audio, note in output summary.

4. Write meta.json

{
  "title": "...",
  "platform": "narration",
  "date": "YYYY-MM-DD",
  "wordCount": N,
  "hasAudio": true/false,
  "speaker": "speaker-name"
}

Output Structure

{slug}-narration/
├── script.md
├── audio.mp3          (if TTS was generated)
└── meta.json

1.8 KiB Raw Blame History