SKILL.md
You are running the Voice Podcast Kit skill.
Goal
- Produce a complete podcast episode with distinct cloned voices for each host/guest, intro/outro music, and smooth transitions between segments.
Ask for
- Episode topic and title.
- List of speakers/characters (names and optional voice sample descriptions).
- Whether you should clone voices from provided audio samples, or use existing voice IDs.
- Episode length target and number of segments (intro, main discussion, listener Q&A, outro).
- Any music preferences (genre, mood, tempo).
Workflow
- Confirm speaker lineup and collect voice samples if cloning is requested:
- If audio files are provided, call voiceclone for each speaker. - If no samples, call voicelist to show available presets and let user choose.
- Draft a segment-by-segment outline with speaker assignments and timing.
- Generate intro/outro music:
- Call generate_music with appropriate mood (upbeat for intro, winding down for outro).
- Write scripts for each segment with clear speaker labels.
- For each spoken segment:
- Call tts with the correct voiceid for each speaker. - Use outputformat "mp3" for smooth editing.
- Optionally add transition sounds or music beds between segments.
- Return a production清单:
- Music files (intro/outro) - Each spoken segment as separate audio files - Full episode script text - Assembly suggestions (timeline order)
Response style
- Keep track of which voice_id maps to which speaker name clearly.
- Provide file paths organized by segment and type.
- Suggest editing workflow but leave final assembly to user.
Notes
- Voice consistency across episodes is a key value proposition—encourage users to save voice IDs for recurring hosts.
- If the user wants a sample before full production, offer to generate just the intro + first 2 minutes as a preview.