📖 DeepWiki Documentation: https://deepwiki.com/donghaozhang/video-agent-claude-skill

AI Content Generation Suite

A comprehensive AI content generation package with multiple providers and services, consolidated into a single installable package.

Production-ready Python package with comprehensive CLI, parallel execution, and enterprise-grade architecture

Demo Video

Click to watch the complete demo of AI Content Generation Suite in action

Available AI Models

73 AI models across 12 categories - showing top picks below. See full models reference for complete list.

Text-to-Image (Top Picks)

Model	Cost	Best For
nano_banana_pro	$0.002	Fast & high-quality
gpt_image_1_5	$0.003	GPT-powered generation
flux_dev	$0.004	Highest quality (12B params)

Text-to-Video (Top Picks)

Model	Cost	Best For
veo3	$2.50-6.00	Google's latest with audio
sora_2	$0.40-1.20	OpenAI quality
kling_3_pro	$0.50-1.12	Latest Kling generation

Image-to-Video (Top Picks)

Model	Cost	Best For
veo_3_1_fast	$1.20	Google's latest i2v
sora_2	$0.40-1.20	OpenAI quality
kling_3_pro	$0.50-1.12	Latest Kling generation

Cost-Saving Tip: Use --mock flag for FREE validation: aicp generate-image --text "test" --mock

See full models reference for all 73 models with pricing.

Latest Release

What's New in v1.0.24

• Native structured output for LLM calls (eliminates JSON parsing fragility)
• Character portrait registry for visual consistency across scenes
• Per-chapter output organization with meaningful file names
• Documentation overhaul: slimmed README, updated all 73 models reference
• CI fix for optional rich dependency in vimax test suite

Installation

From PyPI

pip install video-ai-studio

Binary (no Python required)

Download standalone binaries from GitHub Releases:

# Linux
curl -L https://github.com/donghaozhang/video-agent-skill/releases/download/latest/aicp-linux-x86_64 -o aicp
chmod +x aicp

# macOS (Apple Silicon)
curl -L https://github.com/donghaozhang/video-agent-skill/releases/download/latest/aicp-macos-arm64 -o aicp
chmod +x aicp

# Windows
curl -L https://github.com/donghaozhang/video-agent-skill/releases/download/latest/aicp-windows-x64.exe -o aicp.exe

Development Mode

git clone https://github.com/donghaozhang/video-agent-skill.git
cd video-agent-skill
pip install -e .

API Keys Setup

After installation, configure your API keys:

1. Create a .env file:

   curl -o .env https://raw.githubusercontent.com/donghaozhang/video-agent-skill/main/.env.example
   

2. Add your API keys:

   # Required for most functionality
   FAL_KEY=your_fal_api_key_here
   
   # Optional - add as needed
   GEMINI_API_KEY=your_gemini_api_key_here
   OPENROUTER_API_KEY=your_openrouter_api_key_here
   ELEVENLABS_API_KEY=your_elevenlabs_api_key_here
   

3. Get API keys from:

   • FAL AI: https://fal.ai/dashboard (required for most models)
   • Google Gemini: https://makersuite.google.com/app/apikey
   • OpenRouter: https://openrouter.ai/keys
   • ElevenLabs: https://elevenlabs.io/app/settings

Quick Start

CLI Commands

# List all available AI models
aicp list-models

# Generate image from text
aicp generate-image --text "epic space battle" --model flux_dev

# Create video (text -> image -> video)
aicp create-video --text "serene mountain lake"

# Run custom pipeline from YAML config
aicp run-chain --config config.yaml --input "cyberpunk city"

# Analyze video with AI
aicp analyze-video -i video.mp4

# Generate avatar with lipsync
aicp generate-avatar --audio speech.mp3 --image portrait.jpg

# Transcribe audio
aicp transcribe --input audio.mp3

# Generate image grid
aicp generate-grid --text "mountain landscape" --layout 2x2

# Create example configurations
aicp create-examples

Unix-Style Flags

All commands support machine-readable output for scripting and CI:

# JSON output for piping to jq
aicp list-models --json | jq '.text_to_video[]'

# Quiet mode (suppress non-essential output)
aicp create-video --text "sunset" --quiet

# Read prompt from stdin
echo "cinematic drone shot" | aicp create-video --input -

# Stream progress as JSONL events
aicp run-chain --config pipeline.yaml --stream

# Combine for CI usage
aicp run-chain --config pipeline.yaml --json --quiet | jq -r '.outputs.final.path'

Python API

from ai_content_pipeline.pipeline.manager import AIPipelineManager

# Initialize manager
manager = AIPipelineManager()

# Quick video creation
result = manager.quick_create_video(
    text="serene mountain lake",
    image_model="flux_dev",
    video_model="auto"
)

# Run custom chain
chain = manager.create_chain_from_config("config.yaml")
result = manager.execute_chain(chain, "input text")

Documentation

For detailed guides and references, see the full documentation:

Topic	Link
Setup & Installation	Setup Guide
All 73 AI Models	Models Reference
CLI Commands	CLI Reference
YAML Pipelines	Pipeline Guide
Cost Management	Cost Guide
Architecture	Architecture Overview
Package Structure	Package Reference
Testing	Testing Guide
Python API	API Reference
Troubleshooting	FAQ & Troubleshooting
Contributing	Contributing Guide
Changelog	Version History
Development Guide	CLAUDE.md

Contributing

1. Follow the development patterns in CLAUDE.md
2. Add tests for new features
3. Update documentation as needed
4. Test installation in fresh virtual environment
5. Commit with descriptive messages