| name | video-production |
| description | Activate for video marketing strategy, script writing, video specifications for YouTube/TikTok/Instagram Reels/LinkedIn, production workflows, video optimization, thumbnail creation, and video repurposing. Used by content-creator and social-media-manager agents. |
| license | MIT |
| allowed-tools | ["Bash","Read","Write","Edit"] |
Video Production
Video marketing strategy, scripting, AI video generation, and platform optimization.
When to Use
- Video strategy planning
- Script writing for any format
- Storyboard creation
- AI video generation (Veo 3.1)
- Platform specifications lookup
- Production workflow guidance
- Video optimization
- Thumbnail creation
- Video repurposing
- Caption extraction
Setup
This skill leverages the ai-multimodal skill for Veo/Imagen APIs. Ensure:
export GEMINI_API_KEY="your-key"
pip install google-genai python-dotenv pillow
Quick Start
Generate video: node scripts/generate-video.cjs --prompt "product demo" --template product-demo
Create storyboard: node scripts/create-storyboard.cjs --script "path/to/script.md"
Analyze video: node scripts/analyze-video.cjs --video "path/to/video.mp4"
Optimize for platform: node scripts/optimize-for-platform.cjs --video "path/to/video.mp4" --platform tiktok
Extract captions: node scripts/extract-captions.cjs --video "path/to/video.mp4"
Core Capabilities
Video Types & Platform Specs
Load: references/video-types-specs.md
Art Directions
Load: references/video-art-directions.md
Script Templates
Load: references/script-templates.md
Production Workflow
Load: references/production-workflow.md
Video Optimization
Load: references/video-optimization.md
Veo Prompt Engineering
Load: references/veo-prompt-guide.md
Storyboard Format
Load: references/storyboard-format.md
Audio Directives
Load: references/audio-directive-guide.md
Thumbnail Design
Load: references/thumbnail-design-guide.md
Video SEO
Load: references/video-seo-optimization.md
Quality Review Workflow
Load: references/quality-review-workflow.md
Scripts
| Script | Purpose |
|---|
generate-video.cjs | Veo 3.1 wrapper with templates |
create-storyboard.cjs | Imagen 4 storyboard generator |
analyze-video.cjs | Gemini video understanding |
optimize-for-platform.cjs | FFmpeg aspect ratio converter |
extract-captions.cjs | Caption/transcript extraction |
Templates
| Template | Use Case |
|---|
product-demo.json | Product showcase videos |
explainer.json | Educational/explainer videos |
testimonial.json | Customer testimonial format |
short-form.json | TikTok/Reels/Shorts format |
Quick Reference
| Platform | Ratio | Ideal Length | Max |
|---|
| YouTube | 16:9 | 8-15 min | 12h |
| YouTube Shorts | 9:16 | 30-60 sec | 3 min |
| TikTok | 9:16 | 15-60 sec | 60 min |
| Reels | 9:16 | 15-30 sec | 3 min |
| LinkedIn | 16:9 | 1-2 min | 10 min |
Video Types: Tutorial, Demo, Testimonial, BTS, Short-form, Explainer
Video Pipeline (Frame-Based with Reviews)
Best quality workflow using start/end frames with AI-powered review gates:
1. SCRIPT (Gemini 2.5) → Scene timing with start/end frame descriptions
↓
[REVIEW 1] → Grammar, timing, brand voice, hooks (Gemini text)
↓
2. STORYBOARD → Frame prompts with visual continuity
↓
[REVIEW 2] → Prompt clarity, style consistency, feasibility (Gemini text)
↓
3. VOICEOVER (Gemini TTS) → Generate narration with voice control
↓
4. MUSIC (Lyria) → Generate background track with mood control
↓
[REVIEW 3] → Audio clarity, pronunciation, mood match (Gemini audio)
↓
5. START/END FRAMES (Imagen 4) → Generate keyframes per scene
↓
[REVIEW 4] → Quality, abnormalities, prompt adherence (Gemini vision)
↓
6. VIDEO (Veo 3.1) → Generate video between reference frames
↓
[REVIEW 5] → Motion quality, artifacts, temporal consistency (Gemini video)
↓
7. MIX (FFmpeg) → Combine video + voiceover + music + SFX
↓
[REVIEW 6] → A/V sync, transitions, overall polish (Gemini video)
↓
8. OPTIMIZE → Platform-specific versions
Key: Every step reviewed by AI before proceeding. Catches typos, abnormal actions, AI artifacts.
Review capabilities from ai-multimodal:
- Text analysis: Script grammar, timing validation
- Vision: Frame quality, abnormality detection (extra limbs, distorted faces)
- Video: Motion quality, temporal consistency, artifact detection
- Audio: Speech clarity, pronunciation errors, mood matching
Time: 25-35 min | Cost: $5-15/video + $0.12 reviews
Workflow
Planning
- Define goal and audience
- Choose video type and platform
- Write script using templates
- Create storyboard
Production (AI-Generated)
- Generate prompts from script
- Generate video clips (Veo 3.1)
- Review and iterate
- Assemble scenes
Post-Production
- Optimize for target platform
- Extract/add captions
- Create thumbnail
- Publish
Output Format
## Video Brief: [Title]
**Platform:** [YouTube/TikTok/etc.]
**Type:** [Tutorial/Demo/etc.]
**Length:** [Duration]
**Aspect Ratio:** [16:9/9:16/1:1]
### Script
[Scene-by-scene breakdown]
### Storyboard
[Key frames with timing]
### Veo Prompts
[Per-scene prompts]
### Visual Notes
[Shot descriptions]
Best Practices
- Hook in first 3 seconds
- Design for sound-off (captions)
- Strong thumbnail with faces
- Clear CTA at end
- Repurpose for multiple platforms
- Use trending audio when applicable
- Optimize title/description for SEO
- A/B test thumbnails
Cost Estimation
| Component | Cost |
|---|
| Veo 3.1 (2-4 clips) | $2-5 |
| Imagen 4 storyboard | $0.04-0.20 |
| Gemini API calls | $0.50-1.00 |
| Per video | $3-6 |
Integration
This skill works with:
ai-multimodal - Video generation and analysis
creativity - Creative direction, style templates, color psychology, voiceover styles
media-processing - FFmpeg optimization
content-marketing - Content strategy
social-media - Distribution planning
Resources