# kordoc-korean-document-parser

This skill turns Korean HWP 5.x, HWPX and PDF files into Markdown and structured blocks. Useful when you need text, tables or form fields out of a document.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode · 774 installs · ★ 88
- Author: reason-machines
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/reason-machines/trending-skills/tree/HEAD/skills/kordoc-korean-document-parser/SKILL.md
- Updated: 2026-10-09
- HTML: https://affpapa.org/en/nevibecode/skills/kordoc-korean-document-parser

## What it does

Under the hood it is the kordoc library — a TypeScript package plus a CLI. It reads the proprietary HWP binary format, pulls out tables, recognises form fields, diffs two revisions of a document and can generate HWPX back from Markdown. Format detection runs on magic bytes, and there are dedicated parsers: `parseHwp`, `parseHwpx`, `parsePdf`.

Three ways in. The CLI: `npx kordoc document.hwpx` prints to stdout, `-o` writes a file, `--format json` returns blocks with metadata, and `watch` converts new files as they land in a folder. The programmatic API: `parse` hands back `markdown`, `blocks`, `metadata`, `outline` and `warnings`. Or run it as the `kordoc-mcp` MCP server, which exposes `parse_document`, `compare_documents`, `parse_form` and more.

PDF support needs `pdfjs-dist` installed separately — skip it and PDF parsing simply fails. Scanned pages go through a pluggable OCR: you pass your own async function in the `parse` options, there is no built-in engine. Encrypted files and ZIP bombs come back as an error code instead of an empty result, which is what you want when a batch job runs unattended.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add reason-machines/trending-skills --skill kordoc-korean-document-parser -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/kordoc-korean-document-parser/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/kordoc-korean-document-parser/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add reason-machines/trending-skills --skill kordoc-korean-document-parser -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/kordoc-korean-document-parser/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add reason-machines/trending-skills --skill kordoc-korean-document-parser -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/kordoc-korean-document-parser/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install reason-machines/trending-skills kordoc-korean-document-parser
```

2. **Manually: put the files in place**
- File: `.github/skills/kordoc-korean-document-parser/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add reason-machines/trending-skills --skill kordoc-korean-document-parser -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/reason-machines/trending-skills --path skills/kordoc-korean-document-parser
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/kordoc-korean-document-parser/SKILL.md` — The legacy .windsurf/skills/ path still works

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add reason-machines/trending-skills --skill kordoc-korean-document-parser -a opencode
```

### Claude.ai

**Claude.ai**
- Zip the folder and upload it in Settings → Capabilities → Skills

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/reason-machines/trending-skills/tree/HEAD/skills/kordoc-korean-document-parser/SKILL.md

## When to use it
- convert a Korean HWP file to Markdown you can edit
- pull tables out of HWPX as structured data
- diff two revisions of a document
- collect form field values as JSON
- set up an MCP server so the agent reads HWP and PDF itself

## Example prompts
- parse `document.hwpx` to Markdown and show me the tables
- compare `v1.hwp` and `v2.hwpx`, what changed
- extract the form fields from `report.hwp`

## Similar
- [hwp](https://affpapa.org/en/nevibecode/skills/hwp.md) — skill · 8500 installs
- [Skywork Document](https://affpapa.org/en/nevibecode/skills/skywork-document.md) — skill · 371 installs
- [nutrient-document-processing](https://affpapa.org/en/nevibecode/skills/nutrient-document-processing.md) — skill · 4500 installs
- [business-document-generator](https://affpapa.org/en/nevibecode/skills/business-document-generator.md) — skill · 1400 installs
- [game-design-document](https://affpapa.org/en/nevibecode/skills/game-design-document.md) — skill · 1100 installs

From the same set: [reason-machines/trending-skills](https://affpapa.org/en/nevibecode/by/reason-machines/trending-skills.md) — 49 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md