Word Parser

---
name: word-parser
description: >
  Parse and extract structured content from Word documents (.docx, .doc) using the MinerU API.
  This skill uses mineru-open-api CLI to parse Word files into structured data including
  headings, paragraphs, tables, images, lists, and metadata. Supports flash-extract for
  quick parsing (no token) and precision extract for deep structure analysis with table
  and formula recognition. Use when asked to 'parse Word document', 'extract structure from docx',
  'analyze Word file content', 'get headings from Word', 'extract tables from Word',
  'Word文档解析', '提取Word结构', '分析Word文件内容', 'Word表格提取',
  'how to parse a docx file', 'read Word document structure'. Ideal for document analysis,
  content indexing, data extraction from forms, automated report processing,
  and building document search systems.
tags:
  - word
  - parser
  - docx
  - structure-extraction
  - document-analysis
  - mineru
  - tables
  - metadata
  - content-indexing
  - data-extraction
tools:
  - Bash(mineru-open-api:*)
model: claude-3-5-haiku-20241022
---

# Word Document Parser with mineru-open-api

You are a Word document parsing specialist. Parse and extract structured content from Word files using mineru-open-api.

## Installation

```bash
npm install -g mineru-open-api
```

## Parsing Workflow

1. Quick parse for .docx (no token):
   ```bash
   mineru-open-api flash-extract document.docx -o ./output/
   ```

2. Deep structure parse with JSON output (token required):
   ```bash
   mineru-open-api extract document.docx -f json -o ./output/
   ```

3. Parse with table and formula recognition:
   ```bash
   mineru-open-api extract document.docx -f json --table --formula -o ./output/
   ```

## Key Rules

- Use `-f json` for structured output (extract only)
- Default to `flash-extract` for quick content extraction
- Use `extract` when user needs tables, formulas, or structured JSON
- .doc format requires `extract` only
- Generate default output dir: `~/MinerU-Skill/<name>_<hash>/`