Viral Video Creator

Viral Video Creator

Comment

Viral Marketing Agent

AI-powered system that discovers trending Instagram content and instantly generates satirical, share-worthy videos.

✨ Key Features

  • Content Scraping – Pull top-performing posts from Instagram accounts (primary via Bright Data MCP, fallback via Apify Instagram Scraper).
  • Idea Selection UI – React dashboard lets you browse scraped posts and pick an idea to amplify.
  • AI Video Generation – MiniMax MCP turns the selected idea into an auto-edited video (script, voice, visuals).
  • Download & Share – Preview the result in-browser and export for rapid social distribution.

run FastAPI server from the backend directory

cd backend source .venv/bin/activate # or .venv\Scripts\activate on Windows uvicorn main:app --reload |--------------|------------| | Backend | Python 3.11 (FastAPI) | | Frontend | React 18 + Vite | | AI / MCP | MiniMax MCP (video generation) | | Scraping | Bright Data MCP (primary) • Apify Instagram Scraper (backup) | | Auth / Config| dotenv for secrets |

🚀 Quick Start

1. Clone and prepare

git clone https://github.com/your-org/viral-marketing-agent.git
cd viral-marketing-agent

2. Environment variables

Create a .env file at the project root:

# MiniMax (video generation)
MINIMAX_API_KEY=your_minimax_key

# Bright Data (Instagram scraping)
BRIGHTDATA_API_TOKEN=your_brightdata_token

# Apify (backup scraper)
APIFY_API_TOKEN=your_apify_token

Never commit .env – it is git-ignored.

3. Backend setup

# create virtual env
python -m venv .venv
source .venv/bin/activate  # Windows: .venv\Scripts\activate

# install deps
pip install -r backend/requirements.txt

# run FastAPI server
uvicorn backend.main:app --reload

Server starts at http://localhost:8000.

4. Frontend setup

cd frontend
npm install
npm run dev     # or: bun install && bun run dev

Frontend is now on http://localhost:5173 (proxy to backend is pre-configured).

⚙️ MCP Server Commands

The backend wraps two MCP processes. They start automatically, but you can run them manually for troubleshooting:

# MiniMax
uvx minimax-mcp -y

# Bright Data
npx @brightdata/mcp

📚 Usage Guide

  1. Open the web app.
  2. “Scrape New Ideas” to fetch latest viral posts from Instagram.
  3. Click any tile to view post details, caption & engagement stats.
  4. Press “Generate Video” → backend calls MiniMax MCP.
  5. Watch the preview; click “Download” to save MP4.

Typical turnaround is under 60 seconds.

📝 Development Scripts

# lint & format
ruff check .
black .

# run unit tests
pytest

🛠️ Troubleshooting

ModuleNotFoundError: No module named 'services'

If you see this error when starting the backend it usually means the server is being launched from the project root rather than inside the backend package.

Run the server like so:

cd backend
source .venv/bin/activate
uvicorn main:app --reload

main.py uses relative imports (from services import …) that expect the working directory to be backend/. Running from within that folder (or using python -m backend.main) ensures the imports resolve correctly.

📄 License

MIT

Built With

Share this project:

Updates