Summary
Automates YouTube operations through the YouTube Data API v3, handling video uploads, Shorts publishing, playlist management, and analytics retrieval. Built as part of a fitness influencer automation pipeline but can be used standalone for any YouTube content workflow. Exposes standard YouTube API operations as MCP tools so Claude can upload videos, create playlists, fetch channel stats, and pull engagement metrics without you writing API client code. Comes from a larger dev sandbox repo focused on AI automation assistants, so it follows a directive/orchestration/execution pattern. Reach for this when you want Claude to manage your YouTube channel programmatically, whether for automated content publishing, analytics reporting, or batch playlist operations.
Keep your Mac awake while Claude Code and 40+ AI agents run. Sleeps when they're idle.
One time payment $9 →
Integrate web data into your AI product. One API to scrape website & brand data.
Get API Key Now →
Agent, run crypto. Access onchain data & trade routes via 1inch.
Install now →
On Capafy, your Skill runs online 24/7 as an agent product, and you get paid every time someone uses it.
Start earning →
AI Assistants Hub - Development Sandbox
Central development workspace for AI-powered automation assistants using the 3-Layer Architecture (Directive → Orchestration → Execution).
Quick Navigation
See PROJECT_INDEX.md for detailed file paths and switching between projects.
Active Projects
| Project | Status | Directory | Production URL |
|---|---|---|---|
| Fitness Influencer | Live | projects/fitness-influencer/ | Railway |
| Interview Prep | Live | interview-prep-pptx/ | Railway |
| Amazon Seller | Dev | projects/amazon-seller/ | - |
| Naples Weather | Dev | projects/naples-weather/ | - |
Repository Structure
dev-sandbox/
├── projects/ # Individual AI assistant projects
│ ├── fitness-influencer/ # Fitness content automation
│ │ ├── src/ # Python scripts
│ │ ├── frontend/ # Web interface
│ │ └── README.md
│ ├── interview-prep/ # → symlink to interview-prep-pptx/
│ ├── amazon-seller/ # Amazon SP-API automation
│ │ └── src/
│ ├── naples-weather/ # Weather report generator
│ └── shared/ # Shared utilities across projects
│ ├── ai/ # AI services (Grok)
│ ├── google/ # Google APIs (Gmail, Sheets)
│ ├── analytics/ # Business analytics
│ └── communication/ # SMS, email
│
├── interview-prep-pptx/ # Railway-linked Interview Prep project
│ ├── src/
│ ├── frontend/
│ ├── Procfile
│ └── railway.json
│
├── execution/ # All execution scripts (skill access)
├── directives/ # SOPs in Markdown format
├── .claude/skills/ # Skill configurations
│
├── PROJECT_INDEX.md # Quick navigation guide
├── index.html # Main website homepage
├── setup_form.html # Fitness Influencer setup form
└── deploy_to_skills.py # Deployment pipeline
Working on Projects
Switch to a Project
cd projects/fitness-influencer # Fitness Influencer
cd interview-prep-pptx # Interview Prep
cd projects/amazon-seller # Amazon Seller
Key Locations by Project
| Project | Scripts | Frontend | Skill | Directive |
|---|---|---|---|---|
| Fitness | projects/fitness-influencer/src/ | projects/fitness-influencer/frontend/ | .claude/skills/fitness-influencer-operations/ | directives/fitness_influencer_operations.md |
| Interview | interview-prep-pptx/src/ | interview-prep-pptx/frontend/ | .claude/skills/interview-prep/ | directives/interview_prep.md |
| Amazon | projects/amazon-seller/src/ | TODO | .claude/skills/amazon-seller-operations/ | directives/amazon_seller_operations.md |
Shared Utilities
Common services used across projects (located in projects/shared/):
| Utility | Path | Used By |
|---|---|---|
| Grok AI Images | shared/ai/grok_image_gen.py | Fitness, Interview |
| Gmail Monitor | shared/google/gmail_monitor.py | Fitness, Amazon |
| Revenue Analytics | shared/analytics/revenue_analytics.py | Fitness, Amazon |
| Twilio SMS | shared/communication/twilio_sms.py | Fitness, Amazon |
Deployment
Deploy to Railway
cd interview-prep-pptx
railway up
railway domain
Deploy to Skills
python deploy_to_skills.py --project fitness-influencer-operations
Environment Variables
All projects share the root .env file:
# AI Services (All projects)
ANTHROPIC_API_KEY=xxx
XAI_API_KEY=xxx
# Google APIs (Fitness, Amazon)
GOOGLE_CREDENTIALS_PATH=credentials.json
# Amazon SP-API (Amazon only)
AMAZON_SELLER_ID=xxx
AMAZON_CLIENT_ID=xxx
AMAZON_CLIENT_SECRET=xxx
# Video Services (Fitness only)
SHOTSTACK_API_KEY=xxx
CREATOMATE_API_KEY=xxx
# Communication (Fitness, Amazon)
TWILIO_ACCOUNT_SID=xxx
TWILIO_AUTH_TOKEN=xxx
3-Layer Architecture
- Directives (
directives/) - SOPs defining what to do - Orchestration - Claude reads directives, makes decisions
- Execution (
execution/) - Deterministic Python scripts
Development Workflow
- Edit in project folder -
projects/{project}/src/orinterview-prep-pptx/src/ - Copy to execution - For skill access:
cp {project}/src/*.py execution/ - Update skill -
.claude/skills/{skill-name}/SKILL.md - Deploy - Railway or
deploy_to_skills.py
Jumping Between Projects
When switching projects mid-session, just say:
- "Let's work on Amazon seller" → Files in
projects/amazon-seller/ - "Switch to fitness influencer" → Files in
projects/fitness-influencer/ - "Back to interview prep" → Files in
interview-prep-pptx/
Features can be shared between projects using projects/shared/.
Session History
Featured
Keep your Mac awake while Claude Code and 40+ AI agents run. Sleeps when they're idle.
One time payment $9 →Integrate web data into your AI product. One API to scrape website & brand data.
Get API Key Now →Agent, run crypto. Access onchain data & trade routes via 1inch.
Install now →On Capafy, your Skill runs online 24/7 as an agent product, and you get paid every time someone uses it.
Start earning →Configuration
YOUTUBE_CREDENTIALS_PATHPath to OAuth client secrets JSON from Google Cloud Console
YOUTUBE_TOKEN_PATHPath to store OAuth token (defaults to ~/.google/youtube_token.pickle)
Registryactive
Packageyoutube-creator-mcp
TransportSTDIO
UpdatedJan 16, 2026
