Crawl4AI MCP Server

# CPU-only optimized requirements for Crawl4AI MCP Server # Excludes heavy ML dependencies to reduce image size # Core MCP and FastMCP fastmcp>=0.1.0 pydantic>=2.0.0 asyncio typing-extensions python-dotenv>=1.0.0 # Web crawling essentials (crawl4ai base dependencies without ML) aiosqlite~=0.20 lxml~=5.3 numpy>=1.26.0,<3 pillow>=10.4 playwright>=1.49.0 requests~=2.26 beautifulsoup4~=4.12 tf-playwright-stealth>=1.1.0 xxhash~=3.4 rank-bm25~=0.2 aiofiles>=24.1.0 snowballstemmer~=2.2 pyOpenSSL>=24.3.0 psutil>=6.1.1 nltk>=3.9.1 rich>=13.9.4 cssselect>=1.2.0 httpx>=0.27.2 httpx[http2]>=0.27.2 fake-useragent>=2.0.3 click>=8.1.7 pyperclip>=1.8.2 chardet>=5.2.0 aiohttp>=3.11.11 brotli>=1.1.0 humanize>=4.10.0 lark>=1.2.2 # CPU-only ML libraries (lightweight alternatives) scikit-learn # For basic ML without torch alphashape>=1.3.1 shapely>=2.0.0 # File processing - explicit dependencies to avoid conflicts markitdown>=0.0.1a2 pdfminer-six>=20250506 mammoth>=1.9.1 openpyxl>=3.1.5 python-pptx>=1.0.2 xlrd>=2.0.1 # Search and external APIs googlesearch-python>=1.3.0 # YouTube transcript extraction (youtube-transcript-api v1.1.0+) youtube-transcript-api>=1.1.0 # Note: crawl4ai is NOT included here to avoid its heavy ML dependencies # We'll install it with --no-deps and manually specify what we need