Converts PDFs to Markdown and vice versa using PyMuPDF for extraction and ReportLab for generation. Exposes pdf_to_markdown and markdown_to_pdf tools through MCP, with options for image extraction, page sizing, margins, and font configuration. The engine automatically detects CJK characters and switches between ReportLab, xhtml2pdf, and fpdf2 based on content complexity. Handles Unicode properly across platforms by finding system fonts like Arial Unicode, PingFang SC, or Noto Sans CJK. Reach for this when you need bidirectional PDF conversion with Claude, especially if you're working with Chinese, Japanese, or Korean documents where font handling matters.
claude mcp add --transport stdio huoshuiai42-huoshui-pdf-converter -- uvx huoshui-pdf-converter