video-producer

name	video-producer
description	End-to-end AI video production pipeline. Takes a brief, generates storyboard, creates scene images/clips via fal-ai, adds voiceover (ElevenLabs/OpenAI TTS), burns in captions, mixes background music, and delivers platform-ready video. Use with the Video Producer persona for a Pexo-like conversational video creation experience.
compatibility	Created for Zo Computer
metadata	{"author":"marlandoj.zo.computer","persona":"Video Producer (a610972e-fb54-43f5-8383-a7b91a4fbd42)","version":"2.0.0"}

Video Producer Skill

End-to-end video production from idea to export-ready video.

Quick Start

cd /home/workspace/Skills/video-producer/scripts

# Full pipeline — brief to finished video
bun video-producer.ts produce \
  --brief "15-second TikTok ad for tallow balm, punchy hook, natural ingredients angle" \
  --platform tiktok \
  --output /home/workspace/Projects/ffb/05-video/output.mp4

# Individual commands
bun video-producer.ts storyboard --brief "..." --platform tiktok --output storyboard.json
bun video-producer.ts generate --storyboard storyboard.json --output-dir ./scenes/
bun video-producer.ts voiceover --script "Your skin deserves better." --voice Rachel --output vo.mp3
bun video-producer.ts captions --script script.txt --style tiktok-bold --output captions.ass
bun video-producer.ts assemble --scenes ./scenes/ --audio vo.mp3 --music acoustic-warm --captions captions.ass --output final.mp4
bun video-producer.ts optimize --input final.mp4 --platform tiktok --output final-tiktok.mp4

Commands

Command	Description
storyboard	Generate storyboard JSON from a text brief
generate	Generate scene images and video clips from storyboard
voiceover	Generate voiceover audio from script text
captions	Generate timed ASS/SRT caption file
assemble	Compose final video from scenes + audio + captions
produce	Full pipeline: storyboard → generate → voiceover → captions → assemble
optimize	Re-encode for a specific platform
preview	Quick low-res preview from scene assets
publish	Upload to zo.space and publish to TikTok/Facebook/Instagram via Blotato
calendar	Batch-produce videos from a content calendar markdown file
translate	Translate storyboard text/voiceover to another language (es, fr, pt, de, ja, zh)
variants	Generate A/B variant storyboards with different hooks/music/styles
lipsync	Generate lip-synced talking-head video from face image + audio
analytics	Pull platform metrics and generate improvement recommendations

Run bun video-producer.ts --help or bun video-producer.ts <command> --help for full options.

Dependencies

• fal-ai-media — image generation + image-to-video. Bundled at scripts/vendor/fal-ai-media/; run bun install there once. Falls back to /home/workspace/Skills/fal-ai-media/ if vendored deps are absent. (FAL_KEY required in Zo Secrets.)
• elevenlabs-skill — voiceover generation (ELEVENLABS_API_KEY in Zo Secrets, paid plan required)
• ffmpeg — video composition, pre-installed on Zo
• OpenAI TTS — free voiceover fallback (OPENAI_API_KEY in Zo Secrets)
• OpenAI GPT-4o-mini — translation for translate command (OPENAI_API_KEY)
• Blotato MCP — social media publishing (BLOTATO_API_KEY in env) for publish/analytics commands
• fal.ai sadtalker/wav2lip — lip-sync generation for lipsync command (FAL_KEY)

Platform Presets

Platform	Aspect	Resolution	Max Duration
tiktok	9:16	1080×1920	60s
reels	9:16	1080×1920	90s
youtube-shorts	9:16	1080×1920	60s
facebook	4:5	1080×1350	60s
youtube	16:9	1920×1080	unlimited

Character Integration

Use with ai-character-builder to apply a character's visual style, voice, and personality to videos:

# Create a character first
bun /home/workspace/Skills/ai-character-builder/scripts/generate-identity.ts --name "Flora Belle" --output character.json
bun /home/workspace/Skills/ai-character-builder/scripts/generate-avatar.ts --identity character.json
bun /home/workspace/Skills/ai-character-builder/scripts/generate-voice.ts --identity character.json

# Use character in video production — auto-applies style, voice, prompt enrichment
bun video-producer.ts produce --brief "tallow balm ad" --character character.json --platform tiktok

# Character avatar as lip-sync source
bun video-producer.ts lipsync --character character.json --audio voiceover.mp3

The --character flag is supported on: produce, generate, lipsync

Assets

• assets/music/ — Royalty-free background tracks (upbeat-corporate, chill-lofi, cinematic-epic, acoustic-warm)
• assets/templates/caption-styles/ — ASS subtitle presets (tiktok-bold, minimal-white, brand-ffb)

Cost Estimate

Component	Model	Cost (15s video, 4 scenes)
Scene images	nano-banana-2 @ $0.04/img	$0.16
Image-to-video	kling-v3-std @ $0.07/sec	~$1.12
Voiceover	ElevenLabs ~$0.03	$0.03
Assembly	ffmpeg (local)	$0.00
Total		~$1.31