Zerox for Claude Code provides a powerful bridge between static documents and actionable Markdown text. Leveraging the zerox library and GPT-4o vision, this skill enables developers to extract structured text from scanned PDFs, Word documents, PowerPoint presentations, and images directly within their workspace. It supports both immediate foreground processing for small files and asynchronous background tasks for large, complex documents, ensuring Claude can ingest and analyze content from almost any source while maintaining formatting integrity.
主要功能
01Multi-format support for PDF, DOCX, PPTX, and various image types
0216 GitHub stars
03AI-powered OCR using GPT-4o vision for high-accuracy text extraction
04Progress monitoring and system notifications for long-running conversion tasks
05Background processing mode to handle large or scanned documents without timeouts
06Clean Markdown output optimized for Claude's context and documentation use
使用场景
01Converting legacy PDF documentation into a searchable Markdown knowledge base
02Extracting technical specifications from scanned images for code generation
03Migrating Word-based project requirements into a developer-friendly format