music-to-storyboard

SKILL.md

/music-to-storyboard — Music → Storyboard

Generate a shot-by-shot storyboard from music analysis, with shot types, camera movements, visual descriptions, mood, and transitions. Output is compatible with pull-film's storyboard-generator format.

Usage

/music-to-storyboard <audio_file_or_analysis_json>

Accepts either an audio file (runs analysis first) or a previously saved analysis JSON.

Steps

  1. Validate input (audio file or JSON)
  2. Generate storyboard:
python3 -m music_analyzer storyboard "<input_path>"
  1. Present storyboard as a table or sequence:

    • Shot # with section label and time range
    • Shot type: wide, medium, close-up, etc.
    • Camera movement: static, dolly, orbit, etc.
    • Visual description (Chinese + English)
    • Transition to next shot: cut, dissolve, fade
    • Energy level and mood
  2. Suggest using /storyboard-generator from pull-film to create the visual storyboard

Section → Shot Mapping

Section Default Shot Camera Visual Treatment
Intro Wide → Medium Slow dolly in Environment establishing
Verse Medium, Close-up Static, slow pan Narrative focus
Chorus Wide, Close-up Orbit, fast push Visual climax
Bridge Close-up, Wide Slow dolly out Contrast, reflection
Outro Wide Slow pull-out Closure, fade

Output Format

Compatible with pull-film's storyboard-generator:

{
  "song_title": "example.mp3",
  "total_duration": 240.5,
  "bpm": 128.0,
  "shots": [
    {
      "shot_number": 1,
      "section": "intro",
      "time_range": {"start": 0, "end": 15.5},
      "shot_type": "wide",
      "camera_movement": "slow dolly in",
      "visual_description_zh": "...",
      "visual_description_en": "...",
      "transition": "fade",
      "energy_level": 0.3
    }
  ]
}
Weekly Installs
3
First Seen
Feb 5, 2026
Installed on
opencode2
claude-code2
replit1
amp1
trae1
kimi-cli1