doc-parser

The doc-parser skill provides advanced document parsing and structure extraction capabilities using docling, IBM’s document understanding library. It is designed to process complex documents such as PDFs, Word files, PowerPoint presentations, images, and HTML content while preserving the original document structure. The skill helps users convert unstructured or semi-structured documents into organized formats such as markdown, plain text, or structured dictionaries. It addresses common challenges in document processing, including handling multi-column layouts, extracting embedded tables and figures, performing OCR on scanned files, and maintaining document hierarchy during conversion.

The skill includes capabilities for document parsing, layout analysis, table extraction, OCR processing, and structured content export. It supports multiple document formats including PDF, DOCX, PPTX, HTML, PNG, and JPG files. Users can configure parsing pipelines with options such as OCR enablement and table structure analysis. The skill also supports exporting parsed content into markdown, text, or JSON-like dictionary representations for downstream processing. Integration with MCP tools such as analyze_document_structure and extract_text_from_pdf allows automated workflows for document analysis and extraction.

This skill is useful for researchers, analysts, technical writers, enterprise teams, and developers who need to process large volumes of documents or extract structured information from reports, academic papers, presentations, or scanned files. Common use cases include converting academic papers into markdown, extracting tables from reports, analyzing document layouts, processing scanned PDFs with OCR, and preserving structured content for data pipelines or knowledge management systems. It is particularly valuable for workflows that require accurate document structure retention and automated content extraction.