@hienlh/ppm 0.9.81 → 0.9.82

package/.opencode/skills/video/references/script-create.md

@@ -1,391 +0,0 @@
-Create a production-ready video script with creative direction, AI-ready prompts for visuals/audio, and strategic evaluation.
-
-## Your Brief
-
-<args>$ARGUMENTS</args>
-
-## Video Types
-- `youtube` - Long-form YouTube video (16:9)
-- `reel` - Instagram Reels / TikTok (9:16)
-- `short` - YouTube Shorts (9:16)
-- `explainer` - Explainer (pick platform based on distribution goal)
-- `tutorial` - Tutorial (pick platform based on distribution goal)
-- `testimonial` - Testimonial (pick platform based on distribution goal)
-
-## Required Inputs (ask if missing)
-- **Goal**: what should the viewer do/think/feel after watching?
-- **Audience**: who is this for (role, sophistication, pain level)?
-- **Platform**: YouTube / Shorts / TikTok / Reels / LinkedIn
-- **Length target**: seconds (short-form) or minutes (long-form)
-- **CTA**: subscribe / comment / visit link / start free / book demo
-- **Offer** (if any): what are you promoting (optional)
-- **Art Direction**: use `AskUserQuestion` tool to suggest a few style preferences from video-art-directions reference of `video` skill, include "Feel free to suggest your own style" option.
-
-**IMPORTANT:** Make sure you fully understand the brief and have all necessary context by using `AskUserQuestion` tool to ask 5-8 clarifying questions.
-
-## Workflow
-
-### 1. Parse Arguments & Art Direction Selection
-- Extract video type from `$ARGUMENTS`.
-- Extract topic/subject.
-- Activate `video` skill and load:
-  - `references/video-types-specs.md` - Platform specs
-  - `references/video-art-directions.md` - Art direction options (Top 50)
-- Determine defaults:
-  - Platform, Aspect ratio, Ideal length, Posting times
-- **Select Art Direction**: Based on brand/topic, recommend 1-3 art directions from the reference. Include:
-  - Style category name
-  - Core keywords
-  - Color palette
-  - Background music style
-  - Voiceover style
-
-### 2. Brief + Research (lightweight, fast)
-- Use multiple `researcher` subagents with activated `research` skill for a quick scope definition (Max 3 tool calls per agent).
-- Produce a short **Video Brief**:
-  - Primary promise (what you'll deliver)
-  - 3-5 key points
-  - Likely objections/misconceptions
-  - Unique angle (what's different vs obvious content)
-- Activate `content-marketing` skill to map:
-  - Funnel stage (awareness/consideration/decision)
-  - Content pillar
-  - 1-2 related follow-up videos to create a series
-
-### 3. Brand Alignment
-- If `docs/brand-guidelines.md` doesn't exist, use `AskUserQuestion` to ask some questions to understand the brand guidelines.
-- Activate `brand` skill.
-- If relevant, run `node .opencode/skills/brand/scripts/inject-brand-context.cjs --json` and incorporate:
-  - Tone of voice
-  - Key messaging constraints
-  - Forbidden phrases / claims to avoid
-  - Naming conventions (product name, features)
-
-### 4. Script Creation with AI-Artist Prompts
-- Use `planner` agent to choose appropriate structure.
-- Use `content-creator` agent to draft the script.
-- Use `copywriter` agent to punch up:
-  - Hook (first 3 seconds)
-  - Pattern interrupts
-  - CTA(s)
-- Activate `video` skill and load:
-  - `references/script-templates.md`
-  - `references/audio-directive-guide.md`
-- **Activate `ai-artist` skill** and load:
-  - `references/image-prompting.md` - For Imagen/Veo frame prompts
-  - `references/nano-banana.md` - For Gemini-specific prompting
-  - `references/character-consistency.md` - For character portrait workflow
-- **Activate `ai-multimodal` skill** and load:
-  - `references/character-consistency-portrait-reference.md` - For Nano Banana Pro multi-ref
-- Script requirements:
-  - Hook is explicit and occurs immediately.
-  - For short-form, use **4-scene structure** (Hook → Problem → Solution → CTA) unless brief demands otherwise.
-  - Include **on-screen text** (sound-off friendly).
-  - Include **b-roll / overlay** suggestions.
-  - Include **camera** + **motion** notes.
-  - Keep pacing within guide: <= 6 words/second for VO.
-
-### 5. Character Profiles (if characters present)
-
-If the script features human characters (actors, personas, presenters, mascots):
-
-1. **Define character profiles** with detailed visual descriptions for each unique character:
-   - Name, gender, age range, ethnicity/skin tone
-   - Hair: color, style, length
-   - Eyes: color, shape
-   - Face: shape, distinguishing features
-   - Clothing: specific items, colors, materials
-   - Accessories: glasses, jewelry, etc.
-   - Build: body type
-
-2. **Create identity anchors** - short descriptors to repeat in EVERY scene prompt:
-   ```
-   "{name}: {hair_color} {hair_style} hair, {eye_color} eyes, wearing {clothing}, {accessories}"
-   ```
-
-3. **Add portrait generation prompts** to the script output for use in storyboard phase:
-   ```
-   Portrait Prompt ({name}): Professional portrait photograph of {name}: {full description}.
-   Canon EOS R5, 85mm f/1.4, soft studio lighting, neutral gray background, sharp focus, 4K.
-   ```
-
-**IMPORTANT:** Character portraits will be generated during `/video:storyboard:create` phase using Nano Banana Pro. These portraits serve as identity references for all scene frame generation.
-
-### 6. AI Prompt Generation (ai-artist)
-For each scene, generate production-ready prompts:
-
-**For scenes WITH characters - use Nano Banana Pro with portrait reference:**
-```
-[Identity anchor: "Same person from reference: {identity_anchor}"]
-[Action/Pose in Setting/Environment]
-[Style/Medium from art direction] [Lighting from art direction]
-[Camera/Lens] [Composition] [Quality Modifiers] [Aspect Ratio]
-Maintain EXACT facial features, hair, and clothing from reference.
-```
-
-**For scenes WITHOUT characters - use Nano Banana Flash:**
-```
-[Subject + Details] [Action/Pose] [Setting/Environment]
-[Style/Medium from art direction] [Lighting from art direction]
-[Camera/Lens] [Composition] [Quality Modifiers] [Aspect Ratio]
-```
-
-**Video Prompts (Veo 3.1)**:
-```
-[Start state → End state] [Motion description]
-[Camera movement: pan/tilt/dolly/crane/tracking]
-[Cinematography style from art direction]
-[Scene transitions: cut/fade/dissolve]
-```
-
-**Music Prompts (Lyria)**:
-```
-[Genre + mood from art direction] [BPM range]
-[Intensity curve: intro/build/sustain/resolve]
-[Instruments: specific mentions]
-[Negative: what to avoid]
-```
-
-**SFX Prompts (Veo native)**:
-```
-SFX: {description} as {trigger action} at {timestamp}
-Ambient: {element 1}, {element 2}
-```
-
-### 7. Platform Optimization Pack
-- Use `video` skill references:
-  - `references/video-optimization.md`
-  - `references/video-seo-optimization.md`
-- Generate:
-  - **Title options** (YouTube/LinkedIn)
-  - **Description** (YouTube, includes chapters for long-form)
-  - **Tags** (YouTube)
-  - **Caption** (TikTok/Reels/Shorts)
-  - **Hashtags** (per platform strategy)
-  - **Best posting times** (from platform specs)
-  - **Thumbnail concept** (1-3 variants, text <= 4 words)
-
-### 8. Quality Gate (script-only)
-- Apply Gate 1 from `video/references/quality-review-workflow.md`:
-  - Grammar/typos
-  - Hook strength
-  - CTA clarity
-  - Timing feasibility
-  - Brand voice consistency
-  - **AI prompt completeness**
-- If any check fails, revise once and re-check.
-
-### 9. Strategic Evaluation
-Generate comprehensive evaluation:
-- **Pros**: Strengths of this script approach
-- **Cons**: Potential weaknesses or risks
-- **Recommendations**: Optimization suggestions
-- **Why It Should Work**: Evidence-based reasoning
-- **Challenges**: Production or creative challenges
-- **Unresolved Questions**: Decisions for client/team
-
-## Agents Used
-- `researcher` - Lightweight research/angle discovery
-- `planner` - Structure selection
-- `content-creator` - Script drafting
-- `copywriter` - Hook/CTA polish
-
-## Skills Used
-- `video` - Script templates, pacing, platform specs, art directions
-- `creativity` - Creative direction, style templates, color psychology, voiceover styles
-- `ai-artist` - AI-ready prompts for frames, video, music, SFX
-- `brand` - Voice and messaging constraints
-- `research` - Lightweight research/angle discovery
-- `content-marketing` - Funnel stage + series planning
-- `assets-organizing` - Output path conventions
-
-## Output
-- Script pack → `content/video/{type}/{date}-{slug}/script.md`
-
-## Next Steps
-- Storyboard → `/video:storyboard:create content/video/{type}/{date}-{slug}/script.md`
-
-## Output Format (`script.md`)
-```markdown
----
-title: "{working title}"
-slug: {kebab-case-topic}
-type: {type}
-platform: {platform}
-aspect_ratio: {16:9|9:16|1:1}
-target_length: "{e.g. 45s | 8-12 min}"
-audience: "{who}"
-goal: "{desired outcome}"
-cta: "{cta}"
-art_direction: "{selected style from video-art-directions}"
----
-
-# Video Brief
-
-## Core Promise
-{one sentence}
-
-## Key Points
-- {point}
-- {point}
-- {point}
-
-## Angle
-{what makes this different}
-
----
-
-# Creative Direction
-
-## Art Direction: {style name}
-**Core Keywords:** {keywords from reference}
-**Color Palette:** {hex codes + names}
-**Signature Effects:** {effects from reference}
-
-## Visual Style
-{2-3 sentences describing the overall visual approach}
-
-## Audio Style
-**Music:** {genre, mood, BPM, instruments}
-**Voiceover:** {style, tone, pacing}
-**SFX:** {overall approach to sound design}
-
----
-
-# Characters (if applicable)
-
-## Character Profiles
-
-### {Character Name}
-**Description:** {gender}, {age}, {ethnicity/skin tone}, {hair: color, style, length}, {eye color}, {face shape, features}, {build}
-**Clothing:** {detailed clothing description}
-**Accessories:** {glasses, jewelry, etc.}
-**Identity Anchor:** "{hair_color} {hair_style} hair, {eye_color} eyes, wearing {clothing}, {accessories}"
-
-**Portrait Prompt (Nano Banana Pro):**
-```
-Professional portrait photograph of {name}: {full description}.
-Canon EOS R5, 85mm f/1.4, soft studio lighting, neutral gray background, sharp focus, 4K.
-```
-
----
-
-# Script
-
-## Scene 1: {name}
-**Time:** {MM:SS}-{MM:SS} ({duration}s)
-
-### Visual
-**Shot/Camera:** {wide/medium/close + movement}
-**Start Frame:** {exact visual description}
-**End Frame:** {exact visual description}
-**Motion:** {what changes between frames}
-**B-roll/Overlay:** {ideas}
-
-### AI Prompts
-
-**IMPORTANT**:
-- DO NOT include any color hex codes in the prompts.
-- If characters present: Use Nano Banana Pro with portrait reference. Include identity anchor in prompt.
-- If no characters: Use Nano Banana Flash for speed.
-
-#### Start Frame (Nano Banana Pro if characters, Flash if no characters)
-```
-{If characters: "Same person from reference: [identity anchor]. "}
-{Complete prompt with style, lighting, composition, aspect ratio}
-{If characters: "Maintain EXACT facial features, hair, and clothing from reference."}
-```
-
-#### End Frame (Nano Banana Pro if characters, Flash if no characters)
-```
-{If characters: "Same person from reference: [identity anchor]. "}
-{Complete prompt maintaining visual continuity}
-{If characters: "Maintain EXACT facial features, hair, and clothing from reference."}
-```
-
-#### Video Generation (Veo 3.1)
-```
-{Complete Veo prompt: start→end state, motion, camera, cinematography}
-```
-
-### On-Screen Text
-- {short caption (<= 4-6 words)}
-
-### Audio
-**Voiceover:** "{line}" (TTS: {Voice}, {Style})
-**Music:** {mood/bpm/intensity}
-**SFX:** {timestamp}: {sound description}
-
-#### Music Prompt (Lyria)
-```
-{Complete Lyria prompt: genre, mood, BPM, instruments, intensity curve}
-```
-
----
-
-{Repeat for each scene...}
-
----
-
-# Platform Optimization
-
-## Titles (pick 1)
-1. {title}
-2. {title}
-3. {title}
-
-## YouTube Description (if applicable)
-{template w/ chapters}
-
-## Tags (YouTube)
-- {tag}
-
-## Caption + Hashtags (Short-form)
-{caption}
-
-## Posting Times
-{best days/times}
-
-## Thumbnail Concepts
-- {concept with visual description}
-
----
-
-# Strategic Evaluation
-
-## Pros
-- {strength 1}
-- {strength 2}
-- {strength 3}
-
-## Cons
-- {weakness/risk 1}
-- {weakness/risk 2}
-
-## Recommendations
-1. {actionable recommendation}
-2. {actionable recommendation}
-3. {actionable recommendation}
-
-## Why It Should Work
-{2-3 paragraphs explaining the strategic reasoning: audience alignment, hook effectiveness, platform optimization, competitive differentiation}
-
-## Challenges
-| Challenge | Mitigation |
-|-----------|------------|
-| {challenge 1} | {how to address} |
-| {challenge 2} | {how to address} |
-
-## Unresolved Questions
-- [ ] {question for client/team decision}
-- [ ] {question for client/team decision}
-```
-
-## Examples
-```
-/video:script:create youtube "complete guide to Claude Code"
-/video:script:create reel "3 marketing automation tips"
-/video:script:create explainer "product onboarding flow"
-/video:script:create tutorial "setting up email campaigns"
-/video:script:create short "ClaudeKit intro" --art-direction "Modern Tech Futurism"
-```

package/.opencode/skills/video/references/script-templates.md

@@ -1,107 +0,0 @@
-# Video Script Templates
-
-Scripts with precise timing for AI video generation (Imagen 4 + Veo 3.1).
-
-## Frame-Based Script Structure
-
-Each scene defines START and END frames for Imagen 4 generation:
-
-```markdown
-## Scene {N}: {Title}
-**Time:** {MM:SS}-{MM:SS} ({duration}s)
-
-### Visual
-**Start Frame:** {Exact visual for Imagen 4 - first frame}
-**End Frame:** {Exact visual for Imagen 4 - last frame}
-**Motion:** {What changes between frames}
-**Camera:** {Shot type + movement}
-
-### Audio
-**Voiceover:** "{Script}" (TTS: {Voice}, {Style})
-**Music:** {Prompt: Genre, intensity}
-**SFX:** {Prompt: Sound} at {timestamp}
-
-### Imagen Prompts
-**Start:** {Complete Imagen 4 prompt}
-**End:** {Complete Imagen 4 prompt}
-```
-
-## 8-Second Product Demo
-
-```markdown
-## Scene 1: Hook (0:00-0:02)
-**Start Frame:** Frustrated professional at cluttered desk, head in hands
-**End Frame:** Same person, head lifting, hint of curiosity
-**Motion:** Subtle head movement toward off-screen
-**Camera:** Medium shot, static
-
-**Voiceover:** "Tired of this?" (TTS: Charon, sympathetic)
-**Music:** Tense corporate, subtle | **SFX:** Paper shuffle at 0:00.5
-
-## Scene 2: Solution (0:02-0:04)
-**Start Frame:** Professional upright, smartphone in hand, screen glowing
-**End Frame:** Leaning forward, engaged smile, phone showing app
-**Camera:** Medium shot, slight dolly in
-
-**Voiceover:** "Meet {Product}." (TTS: Charon, confident)
-**Music:** Uplifting, building | **SFX:** Chime at 0:02.2
-
-## Scene 3: Demo (0:04-0:06)
-**Start Frame:** Close-up phone, cluttered interface, finger near button
-**End Frame:** Organized interface, success state
-**Camera:** Close-up, static
-
-**Music:** Full energy | **SFX:** Click at 0:04.3, chime at 0:05.5
-
-## Scene 4: CTA (0:06-0:08)
-**Start Frame:** Professional at clean desk, smiling, holding phone
-**End Frame:** Same with logo/URL overlay
-**Camera:** Medium, slight dolly out
-
-**Voiceover:** "Start free today." (TTS: Charon, inviting)
-```
-
-## Timing Reference
-
-| Duration | Frames (24fps) | Max Words |
-|----------|----------------|-----------|
-| 1s | 24 | 2-3 |
-| 2s | 48 | 4-6 |
-| 4s | 96 | 8-12 |
-| 8s | 192 | 15-24 |
-
-## Frame Requirements
-
-**Start Frame:** Subject position, expression, environment, lighting, composition
-**End Frame:** Final state, changed elements, emotional resolution, transition setup
-
-## Motion Notation
-
-- Head down → Head up = Lift motion
-- Empty → Filled = Appear motion
-- Far → Close = Dolly in
-- Left → Right = Pan motion
-
-## Hook Formulas
-
-| Formula | Duration | Start → End |
-|---------|----------|-------------|
-| "Stop [mistake]" | 1-2s | Bold text → Problem visual |
-| "What if...?" | 2s | Question → Aspiration |
-| "POV: [discovery]" | 2s | First-person → Reveal |
-| "[Number] ways..." | 2s | Number graphic → Grid |
-
-## Best Practices
-
-1. **Specific frames**: "Navy blazer, confident smile" not "person"
-2. **Continuity**: End frame N ≈ Start frame N+1
-3. **Simple motion**: One movement per 2s segment
-4. **Voice timing**: Max 6 words/second
-5. **SFX precision**: Exact timestamps for sync
-6. **Style tags**: Include "professional storyboard frame" in Imagen prompts
-
-## Resources
-
-- [Storyboard Format](./storyboard-format.md) - Frame generation
-- [Veo Prompt Guide](./veo-prompt-guide.md) - Video from frames
-- [Audio Directive Guide](./audio-directive-guide.md) - TTS/music

package/.opencode/skills/video/references/storyboard-create.md

@@ -1,210 +0,0 @@
-Create video storyboards with START/END frame pairs for Nano Banana Flash to Veo 3.1 pipeline.
-
-## Your Brief
-
-<args>$ARGUMENTS</args>
-
-## Instructions
-
-1. **Activate skills**:
-   - Load `video` skill with `references/storyboard-format.md`, `references/video-art-directions.md` and `references/quality-review-workflow.md`
-   - Load `creativity` skill for style templates, color palettes, visual trends
-   - Load `ai-artist` skill with `references/character-consistency.md` for character portrait workflow
-   - Load `ai-multimodal` skill with `references/character-consistency-portrait-reference.md` for Nano Banana Pro multi-ref
-   - Load `assets-organizing` skill for output path conventions
-2. **Parse input**: Script file path OR text prompt describing video concept
-3. **Determine platform**: YouTube (16:9), TikTok/Reels (9:16), or Square (1:1)
-
-## Workflow
-
-### Step 1: Analyze Input
-- If script file: Extract scenes, timing, visual descriptions
-- If prompt: Generate 4-scene structure (Hook → Problem → Solution → CTA)
-
-### Step 2: Generate Character Portraits (if characters present)
-If the script/brief contains human characters (actors, personas, mascots):
-
-1. **Extract characters**: List all unique characters with their visual descriptions
-2. **Generate portrait** for each character using Nano Banana Pro:
-   ```bash
-   .opencode/skills/.venv/bin/python3 .opencode/skills/ai-multimodal/scripts/gemini_batch_process.py \
-     --task generate --model gemini-3-pro-image-preview \
-     --prompt "Professional portrait photograph of [character_name]: [detailed description]. Canon EOS R5, 85mm f/1.4, soft studio lighting, neutral gray background, sharp focus, 4K." \
-     --aspect-ratio 2:3 --size 4K --output assets/storyboards/{slug}/characters/{character_name}-portrait.png
-   ```
-3. **Review portraits** with `ai-multimodal` vision to validate facial features, clothing, accessories match descriptions
-4. **Save portraits** to `assets/storyboards/{slug}/characters/` for reference in all subsequent frame generations
-
-**IMPORTANT:** Character portraits MUST be generated BEFORE any scene frames. These portraits are the identity anchors for all subsequent frame generation.
-
-### Step 3: Generate Storyboard
-For each scene, create:
-- **Start Frame**: Imagen prompt with style tags + validation criteria
-- **End Frame**: Imagen prompt ensuring continuity + validation criteria
-- **Motion**: Veo directive (static, dolly, pan)
-- **Audio**: Voiceover text, music cue, SFX timestamps
-- **Duration**: 2-3 seconds per scene
-
-### Step 4: Generate Frames
-Use `ai-multimodal` skill (referenced `image-generation.md`) to generate START and END frame images.
-
-**If characters are present in the scene:**
-Use Nano Banana Pro (`gemini-3-pro-image-preview`) with character portrait as `--files` reference:
-```bash
-.opencode/skills/.venv/bin/python3 .opencode/skills/ai-multimodal/scripts/gemini_batch_process.py \
-  --task generate --model gemini-3-pro-image-preview \
-  --files assets/storyboards/{slug}/characters/{character_name}-portrait.png \
-  --prompt "Same person from reference image: [identity anchors]. Now [scene description]. Maintain EXACT facial features, hair, clothing. [style tags] [camera/composition]" \
-  --aspect-ratio <ratio> --size 4K --output assets/storyboards/{slug}/scene-{N}-1-start.png
-```
-
-**If no characters in the scene:**
-Use Nano Banana Flash for speed, or Pro for complex scenes.
-
-If there are referenced images in the brief, use `Multi-Image Composition`.
-Make sure the output frames do not contain any hex color codes.
-
-### Step 5: Review Frames
-Use `ai-multimodal` vision skill to analyze the generated frames and validate each frame pair before proceeding.
-**DO NOT** use the default `Read` tool to read the frames unless `ai-multimodal` skill fails.
-If there are any unacceptable issues: use `ai-multimodal` skill to modify or re-generate and re-check.
-
-**Unacceptable issues:**
-- Typo errors
-- Hex color codes in the frames
-- Duplicate text
-- Blurry text
-- Inconsistent style
-- Meaningless content
-- Unintended motion
-- Unintended continuity
-- Unintended cropping
-
-## Output Structure
-
-Create directory: `assets/storyboards/{date}-{slug}/` containing:
-- `storyboard.md` - Full storyboard document
-- `storyboard.json` - Machine-readable scene data
-- `characters/` - Character portrait images (if characters present)
-  - `{character-name}-portrait.png` - Reference portraits for each character
-- `scene-{N}-1-start.png` - Start frame images
-- `scene-{N}-2-end.png` - End frame images
-
-## Output Format (`storyboard.md`)
-
-```markdown
----
-title: "{working title}"
-slug: {kebab-case-topic}
-type: {type}
-platform: {platform}
-aspect_ratio: {16:9|9:16|1:1}
-target_length: "{e.g. 45s | 8-12 min}"
-audience: "{who}"
-goal: "{desired outcome}"
-cta: "{cta}"
-art_direction: "{selected style from video-art-directions}"
-script: "{script file path or prompt}"
----
-
-# Storyboard: {Title}
-
-**Generated:** {date} | **Aspect:** {ratio} | **Duration:** {total}s
-
----
-
-## Characters (if applicable)
-
-| Character | Portrait | Description |
-|-----------|----------|-------------|
-| {name} | ![{name}](./characters/{name}-portrait.png) | {hair, eyes, clothing, accessories} |
-
-**Identity Anchors** (repeat in every scene prompt):
-- {name}: "{hair_color} {hair_style} hair, {eye_color} eyes, wearing {clothing}, {accessories}"
-
----
-
-## Script
-{script overview}
-{script file path}
-
-### Core Promise
-{one sentence}
-
-### Key Points
-- {point}
-- {point}
-- {point}
-
-### Angle
-{what makes this different}
-
----
-
-## Creative Direction
-
-### Art Direction: {style name}
-**Core Keywords:** {keywords from reference}
-**Color Palette:** {hex codes + names}
-**Signature Effects:** {effects from reference}
-
-### Visual Style
-{2-3 sentences describing the overall visual approach}
-
-### Audio Style
-**Music:** {genre, mood, BPM, instruments}
-**Voiceover:** {style, tone, pacing}
-**SFX:** {overall approach to sound design}
-
----
-
-## Scene {N}: {Name}
-
-| Property | Value |
-|----------|-------|
-| Timing | {start}-{end} |
-| Duration | {N}s |
-| Shot | {wide/medium/close-up} |
-| Motion | {veo directive} |
-
-### Start Frame
-**Prompt:** {imagen prompt with style tags}
-![Start](./scene-{N}-1-start.png)
-
-### End Frame
-**Prompt:** {imagen prompt with style tags}
-![End](./scene-{N}-2-end.png)
-
-### Video Prompts (Veo 3.1):
-\`\`\`
-[Start state → End state] [Motion description]
-[Camera movement: pan/tilt/dolly/crane/tracking]
-[Cinematography style from art direction]
-[Scene transitions: cut/fade/dissolve]
-\`\`\`
-
-### Audio
-- **VO:** "{voiceover text}"
-- **Music:** {mood, bpm}
-- **SFX:** {timestamp: sound}
-
-#### Music Prompt (Lyria)
-```
-{Complete Lyria prompt: genre, mood, BPM, instruments, intensity curve}
-```
-
-### Review Notes
-{validation criteria, continuity checks}
-
----
-```
-
-## Example Usage
-
-```bash
-# From script file
-/video:storyboard:create scripts/product-demo.md
-
-# From prompt
-/video:storyboard:create "30-second TikTok ad for fitness app showing before/after transformation"
-```