From ai-film-production
This skill should be used when the user asks to "convert a storyboard to video prompts", "turn a script into AI video", "plan a video sequence", "create a shot list", "build an AI film timeline", "break down a scene for video generation", or mentions storyboard conversion, shot-by-shot planning, or multi-scene AI video production workflows.
How this skill is triggered — by the user, by Claude, or both
Slash command
/ai-film-production:ai-storyboard-to-videoThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
A structured pipeline for converting written storyboards, scripts, or scene descriptions into complete AI video production packages. Each scene becomes a self-contained prompt set — start frame, animation, camera, and editing instructions — ready for any AI video generation tool.
A structured pipeline for converting written storyboards, scripts, or scene descriptions into complete AI video production packages. Each scene becomes a self-contained prompt set — start frame, animation, camera, and editing instructions — ready for any AI video generation tool.
This skill bridges the gap between narrative writing and technical video generation prompts.
STORYBOARD → SCENE BREAKDOWN → SHOT LIST → PROMPT PACKAGE → EDITING TIMELINE
Parse the storyboard into discrete scenes. Each scene is one continuous location and time block.
Scene boundary triggers:
SCENE [N]
Location: [Where]
Time: [When — time of day, weather, season]
Characters: [Who is present — reference character sheets]
Emotion: [Core emotional tone]
Duration: [Target clip length in seconds]
Summary: [One-sentence description of what happens]
Break each scene into individual shots. Follow the one action per shot rule.
Each shot defines:
| Element | Description |
|---|---|
| Shot # | Sequential number within the scene |
| Type | Wide / Medium / Close-up / Extreme close-up / Over-the-shoulder |
| Subject | Who or what is in frame |
| Action | Single specific action |
| Camera | Movement type, lens, angle |
| Duration | Target seconds for this clip |
Shot type progression — follow cinematic convention:
For each shot, generate the complete prompt set:
## Shot [N] — [Brief Label]
### Start Frame Prompt
[Full image generation prompt with lighting, lens, composition, character attributes]
### Animation Prompt
[Specific action and camera movement description]
### Camera Settings
[Movement type, lens, angle, shake level, speed]
### Duration
[Target seconds]
### Editing Notes
[Speed ramps, transition to next shot, audio cues]
After all shots are generated, produce the assembly timeline:
## Editing Timeline
| Shot | Duration | Transition | Speed | Audio |
|------|----------|-----------|-------|-------|
| 1 | 3s | Cut | 1x | Ambient city |
| 2 | 4s | Cut | 1x | Footsteps |
| 3 | 2s | Speed ramp| 0.5x→1x | Music swell |
| ... | ... | ... | ... | ... |
Total runtime: [sum]s
When converting a storyboard, always return the complete package:
## Production Overview
[Title, genre, target runtime, character sheets referenced]
## Scene Breakdown
[All scenes with location, time, characters, emotion, duration]
## Shot List with Prompts
[Per-scene, per-shot: start frame + animation + camera + editing notes]
## Editing Timeline
[Full assembly table with transitions, speed, and audio]
## Production Notes
[Technical recommendations, generation order, consistency reminders]
ai-character-sheet-generator| Content | Recommended Duration |
|---|---|
| Establishing wide shot | 2-4 seconds |
| Action shot | 3-5 seconds |
| Dialogue reaction | 2-3 seconds |
| Slow-motion detail | 2-3 seconds (at 0.5x) |
| Transition/cutaway | 1-2 seconds |
Keep individual clips short. AI video tools produce cleaner results in 3-5 second clips than in 10+ second clips.
| Mistake | Fix |
|---|---|
| No establishing shot | Always open a new scene with a wide shot |
| Overloading a single shot with multiple actions | One action per shot — edit together in post |
| Inconsistent lighting across shots in same scene | Lock time-of-day and light direction per scene |
| Missing character sheet references in prompts | Paste locked attributes into every start frame prompt |
| No editing timeline | Always produce the assembly table — generation without editing plan wastes effort |
| Clips too long (8-10+ seconds) | Target 3-5 seconds per clip for best quality |
This skill works as the final stage in a three-skill pipeline:
ai-character-sheet-generator → ai-cinematic-video-director → ai-storyboard-to-video
(characters) (shot craft) (full production)
For genre-specific templates, multi-character scene blocking, and parallel action editing, consult references/production-templates.md.
npx claudepluginhub adityaraj0421/ai-cinematic-video-director-claude-skillSplits scripts/stories into scene-by-scene video generation prompts with structured segments (characters, props, scene, sound, mood, shots) for Seedance 2.0, Kling, and Jimeng.
Generates structured text storyboards for video production. Walks through video type, theme, and duration, then outputs scene-by-scene descriptions with timing, visuals, and sound effects.
Turns natural-language ideas into storyboards and generates images, video clips, and audio automatically. Useful for producing brand videos, short films, social reels, or product ads.