Viral Content Engine

Automated content generation pipeline that scrapes viral tweets and generates original content using Claude AI.

Core thesis: Going viral is about output volume, not ad spend or video quality. Post more, clip more, repurpose more.

What It Does

1. Scrapes tweets from top creators using Playwright (headless browser)
2. Scrapes trending topics from X for timely hooks
3. Analyzes engagement metrics and viral patterns
4. Generates three types of original content daily using Claude AI:
   • 10 tweets — punchy, contrarian, action-oriented
   • 1 blog article — 800-1200 words, SEO-friendly
   • 5 reel/TikTok scripts — 30-60 second HOOK/BODY/CTA format
5. A/B variations — optionally generates 3 versions of each tweet (original + 2 hooks)
6. Serves a web dashboard to browse, compare, and copy generated content

Installation

npm install -g viral-content-engine

Prerequisites

• Node.js 18+
• An Anthropic API key
• (Optional) Twitter/X credentials for scraping

Setup

# 1. Create config file
viral-engine init

# 2. Add your API key
# Edit ~/.viral-engine/.env and set ANTHROPIC_API_KEY

# 3. Install Playwright browsers
npx playwright install chromium

Usage

# Full pipeline (scrape + trends + generate)
viral-engine daily

# Full pipeline with A/B tweet variations
viral-engine daily --ab

# Scrape tweets from configured creators
viral-engine scrape

# Scrape trending topics from X
viral-engine trends

# Generate content from latest scraped data
viral-engine generate

# Generate content with A/B tweet variations
viral-engine generate --ab

# Launch web dashboard at localhost:3333
viral-engine view

# Show help
viral-engine --help

Creator Management

Manage which Twitter/X creators are scraped — no need to edit source code.

# Add a creator
viral-engine follow @garyvee

# Remove a creator
viral-engine unfollow @garyvee

# List all followed creators
viral-engine following

Creators are stored in ~/.viral-engine/creators.json. On first run, defaults to a built-in starter list.

A/B Tweet Variations

The --ab flag runs a second pass that generates two alternative versions of each tweet with different hooks and formats. Use it to pick the strongest version before posting.

viral-engine generate --ab

Variations are saved to:

• output/tweets/YYYY-MM-DD_ab-variations.md (readable markdown)
• data/ab-variations/YYYY-MM-DD.json (structured data for the dashboard)

View them side-by-side in the web dashboard under the A/B Tests tab.

Trending Topics

The trends command scrapes the X Explore page for up to 30 trending topics. These are automatically woven into tweet, blog, and reel generation as timely hooks.

viral-engine trends

The daily command runs trends automatically between scraping and generation. If trending scrape fails, the pipeline continues without it.

Configuration

Environment Variables

Set these in ~/.viral-engine/.env or .env in your working directory:

Variable	Required	Description
ANTHROPIC_API_KEY	Yes	Your Claude API key
TWITTER_USERNAME	No	X/Twitter username for scraping
TWITTER_PASSWORD	No	X/Twitter password for scraping
CLAUDE_MODEL	No	Model to use (default: claude-sonnet-4-6)

Web Dashboard

Run viral-engine view to start the dashboard at http://localhost:3333.

Features:

• Browse generated tweets, blogs, and reel scripts
• A/B Tests tab — 3-column grid comparing Original / Variation A / Variation B
• Copy any piece of content to clipboard
• Dark theme with organized tabs
• Markdown rendering

Output Structure

Generated content is saved as markdown files:

~/.viral-engine/
├── data/
│   ├── scraped/          # Raw scraped tweet data (JSON)
│   ├── trends/           # Trending topics by date (JSON)
│   ├── ab-variations/    # A/B variation data by date (JSON)
│   └── history.json      # Pipeline run history
├── output/
│   ├── tweets/           # YYYY-MM-DD.md + YYYY-MM-DD_ab-variations.md
│   ├── blogs/            # YYYY-MM-DD.md
│   └── reels/            # YYYY-MM-DD.md
├── cookies/              # Saved browser sessions
└── creators.json         # Your followed creator list

Troubleshooting

"ANTHROPIC_API_KEY is not set" Run viral-engine init and add your key to ~/.viral-engine/.env.

"Browser not found" / Playwright errors Run npx playwright install chromium to install the browser.

Twitter scraping fails

• Check your credentials in .env
• Twitter may ask for verification — check the logs
• Cookies expire; delete ~/.viral-engine/cookies/ to force re-login

Rate limiting The engine retries Claude API calls automatically with exponential backoff (up to 3 attempts).

Development

# Clone the repo
git clone https://github.com/IntownNFT/viral-content-engine.git
cd viral-content-engine
npm install

# Run in dev mode
npm run daily

# Build
npm run build

License

MIT — Dylan Worrall