What it does This is orchestration glue: a Node.js scheduler that chains AI calls (OpenAI, Gemini, or Claude via Ollama) into YouTube uploads. It generates scripts, thumbnails, titles, and metadata, then pushes them through the YouTube Data API on a fixed timetable. A local dashboard runs on port 3456 for manual triggers and schedule checks.

The interesting bit The README is unusually honest about being a “content farm” enabler — it lists use cases like “Top 10” compilations, kids content, and meditation videos, and claims a “story channel” hit 1M views/month on autopilot. The cost math is the hook: free tier Gemini (60 req/min) plus the YouTube API’s 10,000 daily units means you can theoretically run a channel for $0.

Key highlights

• Swappable AI backends: OpenAI for quality, Gemini for free tier, or local models
• Hardcoded agent pipeline: strategy → script → thumbnail → SEO → publish → analytics
• Setup wizard and .env configuration; targets non-coders
• Runs locally, on a Pi, or cheap VPS; includes cron-like scheduling
• MIT licensed, 1,243 stars

Caveats

• The “success stories” (50K subscribers in 3 months, $5K/month revenue) are unsourced claims in the README
• Thumbnail “A/B testing” and “competitor research” are described aspirationally — no detail on how either actually works
• YouTube’s ToS around automated content is a known risk area; the FAQ hand-waves this with “as long as you create original content”
• No actual video generation shown; the system appears to produce scripts and metadata, not rendered video

Verdict Worth a look if you’re experimenting with AI-to-API pipelines or running low-stakes YouTube experiments. Avoid if you care about platform risk, content quality, or need proof the success stories are real.