{"id":"4ca9c261-88e7-4cc0-9cbc-119954a4030b","slug":"youtube-automation-agent","kind":"mcp","name":"youtube-automation-agent","summary":"Universal Multi-Agent YouTube Production Engine. Seamless Google Flow, Veo 3, Diffusion Studio MCP, and Headless YouTube API automation for Antigravity, Claude Code, Cursor, and Codex.","owner":"isaim0011","repoUrl":"https://github.com/isaim0011/youtube-automation-agent","license":"MIT","licenseTier":"permissive","trust":"community","stars":1,"match":null,"fewStars":true,"format":"mcp","createdAt":"2026-09-24T09:18:47.000Z","addedAt":"2026-10-21T05:10:57.711Z","logoUrl":"https://avatars.githubusercontent.com/u/297376048?u=252f819e252768022073e04d88198c0cf8e7281f&v=4","linkOnly":false,"collection":null,"skillPath":null,"categories":["design-media","developer-tools","browser-web"],"tags":["ai-agent","antigravity","channel-automation","claude-code","cursor","diffusion-studio","google-flow","lofi-music"],"runtime":"binary","local":true,"cloudCapable":false,"cloudCapability":"unknown","cloudReadyOnPlatform":false,"needsApiKey":false,"reviewed":false,"syncedAt":"2026-10-27T01:44:57.526Z","sample":false,"url":"https://conduits.dev/tools/youtube-automation-agent","language":"Python","transports":["stdio"],"install":{},"requirements":{"env":[],"upstreamMayCharge":false},"tools":null,"readme":"# Universal Autonomous YouTube Production Engine & Multi-Agent Growth Framework\n\n[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE)\n[![Platform: Cross-Platform](https://img.shields.io/badge/Platform-Windows%20%7C%20macOS%20%7C%20Linux-brightgreen.svg)]()\n[![Multi-Agent: Universal](https://img.shields.io/badge/Agents-Antigravity%20%7C%20Claude%20Code%20%7C%20Cursor%20%7C%20Codex%20%7C%20Copilot%20%7C%20Aider-purple.svg)]()\n[![Pipeline: Flow%20Studio%20+%20DiffusionStudio%20+%20YouTube%20API](https://img.shields.io/badge/Stack-Google%20Flow%20%7C%20Veo%20%7C%20DiffusionStudio%20%7C%20Go-orange.svg)]()\n\n> **The Open Universal Standard for Autonomous YouTube Channel Scaling.**  \n> An algorithmic, statistical, and multi-agent system designed for **Antigravity, Claude Code, Cursor, Copilot, Codex, Aider**, and custom LLM agents. Automates the full lifecycle: Google Flow & Veo 3 generation, character/scene consistency, Diffusion Studio agentic timeline editing, high-CTR thumbnail synthesis, and zero-touch headless YouTube API uploads.\n\n---\n\n## 🌟 Universal Multi-Agent Compatibility\n\nThis repository is built from the ground up as a **universal agent standard**. It works out of the box with any autonomous coding assistant:\n\n| AI Agent System | Integration Method | Configuration File |\n| :--- | :--- | :--- |\n| **Google Antigravity** | Native Skill Module | `skills/youtube-agent/SKILL.md` |\n| **Claude Code** | Global Agent Instructions / Tool Calling | `AGENT.md` & `CLAUDE.md` |\n| **Cursor & Windsurf** | System Rules & Context Integration | `.cursorrules` & `AGENT.md` |\n| **GitHub Copilot / Workspace** | Instruction Directives | `.github/copilot-instructions.md` |\n| **Aider / Codex / OpenDevin** | Universal System Prompt | `AGENT.md` |\n\n---\n\n## 🚀 Key Features\n\n* **🧠 Universal `youtube-agent` Skill:** Plug-and-play agent blueprint (`AGENT.md`, `SKILL.md`, `.cursorrules`) that executes mathematical CTR scoring (`P(Click)`), retention loop optimization, and reverse-engineered chapter timestamps.\n* **✨ Google Flow & Veo Automation Suite:** Direct companion automation for batch generating photorealistic 1080p video (Veo 3/3.1) and 2K/4K thumbnails (Nano Banana Pro) with automated character/scene consistency.\n* **🎨 Diffusion Studio Agentic Video Editing:** Replaces fragile manual video manipulations with **Diffusion Studio** (`diffusionstudio/editor`, 3k+ ⭐) via Model Context Protocol (MCP) and programmatic JSX compositions.\n* **🎥 Multi-Angle Video Sequencing Engine:** Intelligent video stitching eliminating repetitive 5-second loop fatigue. Cycles multi-shot cinematic footage with progressive color enhancement (`contrast=1.08:saturation=1.15`) and studio fades.\n* **🎵 Broadcast-Grade Audio Concat:** Lossless WAV-to-256kbps AAC audio mastering, loudness normalization, and micro-crossfade transitions.\n* **🖼️ Dual-Variant High-CTR Thumbnails:** Automated generation of both clean photographic immersion (Bilibili/organic) and contrast-badged text graphics (YouTube mobile feed).\n* **⚡ Headless YouTube API Pipeline:** Integration with `porjo/youtubeuploader` utilizing persistent OAuth tokens and structured JSON schema for zero-touch uploads.\n* **🌐 Guided Multi-Platform Social Onboarding:** Built-in agent workflow guiding users step-by-step through setting up and verifying all 4 major platforms (YouTube Shorts, TikTok, Instagram Reels, Bilibili) with synchronized handles, avatars, and bios before entering production.\n* **🗂️ Clean Multi-Channel Isolation:** Enforces clean separation (`channels/<Channel>/projects/<ID>/`) to prevent asset collisions across single or multi-channel networks.\n\n---\n\n## 📂 Repository Architecture\n\n```text\n├── AGENT.md                       # Universal multi-agent operational manual (All LLMs)\n├── .cursorrules                   # Cursor, Windsurf & VS Code AI configuration\n├── skills/\n│   └── youtube-agent/\n│       └── SKILL.md               # Antigravity & Agent skill definition\n├── scripts/\n│   ├── assemble_album_video.py    # Multi-clip sequencing & color-grading engine\n│   ├── generate_thumbnails.py     # Dual-variant 1280x720 thumbnail generator\n│   └── build_metadata.py          # Reverse-engineered SEO & chapter JSON generator\n├── templates/\n│   ├── video_metadata.template.json # Strict API schema for porjo/youtubeuploader\n│   └── client_secrets.example.json  # OAuth 2.0 configuration blueprint\n├── uploader/\n│   └── README.md                  # YouTube Data API v3 setup & headless workflow\n├── .gitignore\n├── LICENSE\n└── README.md\n```\n\n---\n\n## 🛠️ Prerequisites & One-Time Setup\n\n### 1. Google Flow Studio & Veo Automation Extension\nTo generate video clips and images with consistent characters and prompts:\n1. **Login to Google Flow Studio:**\n   * Open your browser and navigate to [flow.google.com](https://flow.google.com/) (or [labs.google](https://labs.google/)).\n   * Sign in with your Google account. Ensure you are active in the workspace where you want media created.\n2. **Install the Veo Automation Extension:**\n   * Install the official companion extension:  \n     👉 **[Veo Automation: VEO & Nano Banana Google Flow](https://chromewebstore.google.com/detail/veo-automation-veo-nano-b/lnadnaegfljibehcgclljdefggflfpkn)** *(Chrome Web Store ID: `lnadnaegfljibehcgclljdefggflfpkn`)*  \n     *(Official Site: [veoautomation.com](https://veoautomation.com/))*\n3. **Agent Protocol:**\n   * Any AI agent will prompt the user to confirm their Google Flow login.\n   * The agent exports formatted prompt batches and instructs the extension to auto-save directly to `channels/<Channel>/projects/<ID>/raw_assets/`.\n\n---\n\n### 2. Google Cloud Console (YouTube Data API v3)\nTo enable automated headless uploads, configure these three settings once in your Google Cloud Project:\n1. **Enable YouTube Data API v3:**\n   * Go to [Google Cloud Console API Library](https://console.cloud.google.com/apis/library/youtube.googleapis.com) and click **ENABLE**.\n2. **Add Test User (Testing Mode Bypass):**\n   * Go to [OAuth Consent Screen](https://console.cloud.google.com/apis/credentials/consent) ➔ **Test Users** ➔ **Add Users** ➔ Enter your YouTube account email.\n3. **Authorized Redirect URI:**\n   * In [Credentials](https://console.cloud.google.com/apis/credentials) ➔ Click your OAuth 2.0 Client ID ➔ Add `http://localhost:8080/oauth2callback`.\n   * Download the JSON file and save it as `client_secrets.json`.\n\n---\n\n## 🚀 Quick Start Guide\n\n### Step 1: Clone Repository\n```bash\ngit clone https://github.com/isaim0011/youtube-automation-agent.git\ncd youtube-automation-agent\n```\n\n### Step 2: Build Album Video\nPlace your raw video loops in `inputs/videos/` and audio WAVs in `inputs/audio/`:\n```bash\npython scripts/assemble_album_video.py\n```\n\n### Step 3: Generate High-CTR Thumbnails\n```bash\npython scripts/generate_thumbnails.py --input raw_photo.jpg --title \"SUNSET LOUNGE\" --sub \"DEEP ORGANIC HOUSE\"\n```\n\n### Step 4: Automated Upload to YouTube\n```bash\n# First run prompts a one-time browser login, creating request.token.\n# All subsequent uploads run completely headless!\n.\\uploader\\youtubeuploader.exe -cache uploader/request.token -secrets uploader/client_secrets.json -metaJSON metadata.json -filename \"output_video.mp4\" -thumbnail \"thumbnail.jpg\"\n```\n\n---\n\n## 📊 Algorithmic Viability Formulas\n\nThe agent enforces mathematical criteria before approving concepts for render:\n\n### 1. Algorithmic Viability Score (AVS)\n\n```math\nAVS = (0.35 \\times CTR_{est}) + (0.40 \\times RET_{est}) + (0.25 \\times SV_{norm})\n```\n\n> **Threshold:** Only concepts with `AVS >= 78.0` proceed to production.\n\n### 2. Competition-to-Demand Ratio (CDR)\n\n```math\nCDR = \\frac{\\text{Monthly Search and Browse Velocity}}{\\text{Active Competitors in Past 60 Days}}\n```\n\n> **Target:** `CDR > 1.5` indicates a high-probability search cluster.\n\n---\n\n## ⚖️ License & Attribution\n\n* Licensed under the [MIT License](LICENSE).\n* Core uploader engine based on [`porjo/youtubeuploader`](https://github.com/porjo/youtubeuploader).\n* Compositions compatible with [`diffusionstudio/editor`](https://github.com/diffusionstudio/editor).\n"}