Facebook Ads Library MCP
MCP server using FastMCP with modular architecture: mcp_server.py exposes tools interfacing with 3 services - scrapecreators_service.py (Facebook Ad Library API), media_cache_service.py (SQLite caching), & gemini_service.py (AI video analysis). Features intelligent batch processing for multi-brand searches & video analysis (~88% token savings), graceful API rate limiting, secure env config via python-dotenv, & persistent caching to avoid redundant calls.