Summary
This connects to TikTok's API to automate creator workflows. You get operations for posting videos programmatically, pulling analytics data, and managing comment moderation. The source repository shows it's part of a larger dev sandbox with multiple automation projects using a three-layer architecture pattern, though the TikTok-specific implementation details aren't visible in the shared codebase structure. Reach for this when you're building content pipelines that need to push to TikTok without manual uploads, or when you want to pull engagement metrics into dashboards or reporting tools. The stdio transport means it runs as a local process that Claude can call directly during conversations.
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
TIKTOK_CLIENT_KEYTikTok app client key
TIKTOK_CLIENT_SECRETTikTok app client secret
TIKTOK_ACCESS_TOKENTikTok user access token
TIKTOK_REFRESH_TOKENTikTok refresh token for token renewal
Registryactive
Packagetiktok-creator-mcp
TransportSTDIO
UpdatedFeb 9, 2026
