Back to Skills

hwp

Use kordoc for agent-native HWP/HWPX document parsing, JSON extraction, diffing, form-field extraction, and Markdown→HWPX reverse conversion (read/convert only — for binary editing use rhwp-edit).

5,188stars561forksUpdated 5/26/2026

Security Assessment

Safe(100/100)
Security Score100/100

About hwp

The hwp skill provides agent-native parsing and conversion workflows for Korean HWP-related document formats using the kordoc engine. It is designed for read-only document analysis and transformation tasks involving `.hwp`, `.hwpx`, and `.hwpml` files. The skill converts documents into AI-friendly Markdown or structured JSON formats so that automated systems, language models, and downstream workflows can process official documents, reports, forms, and structured content more effectively. It also supports document comparison, form-field extraction, page-range parsing, and Markdown-to-HWPX reverse conversion. The skill explicitly focuses on conversion and inspection workflows rather than binary editing or live UI automation.

Its primary capabilities include HWP/HWPX/HWPML to Markdown conversion, structured JSON extraction with metadata and block-level content, batch conversion of multiple documents, page-specific parsing, and extraction of complex structures such as tables, images, and form fields. The skill consistently routes all supported operations through kordoc, either through the CLI or optional Node.js APIs such as `parse()`, `compare()`, `extractFormFields()`, and `markdownToHwpx()`. It also supports directory watch workflows for automated conversion pipelines and reverse conversion from Markdown back into HWPX documents. The environment requires Node.js 18+ and installation of both `kordoc` and `pdfjs-dist`.

This skill is intended for AI automation engineers, document-processing pipelines, Korean enterprise workflows, and developers handling public-sector or business documents in HWP formats. Typical use cases include converting official Korean documents into machine-readable formats, extracting structured data for AI analysis, comparing document revisions, processing application forms, and automating large-scale document transformation tasks. It is especially useful in environments where HWP and HWPX formats are commonly used and where structured extraction is more important than direct document editing.

FAQ

What document formats are supported by the hwp skill?

The skill supports `.hwp`, `.hwpx`, and `.hwpml` document formats for parsing, conversion, comparison, and structured extraction workflows.

Can this skill edit HWP document content directly?

No. The skill is designed for read-only conversion and extraction tasks. Direct binary editing, text replacement, table modification, or structural editing should be handled by the separate `rhwp-edit` skill.

What output formats can be generated?

The skill can generate Markdown, structured JSON, and HWPX output depending on the selected workflow and conversion direction.

Are there any runtime requirements?

Yes. The environment requires Node.js 18+ and installation of both `kordoc` and `pdfjs-dist`. The current kordoc CLI depends on `pdfjs-dist` even when PDF processing is not used.

Can the skill process multiple documents automatically?

Yes. It supports batch conversion of folders or glob paths, and also includes a watch mode for monitoring directories and automatically processing incoming documents.

Install hwp

Download and extract the skill files to your .claude/skills/ directory.

Quick Setup:

  1. Copy the skill folder to .claude/skills/
  2. Claude will automatically detect and use the skill