A high-precision document extraction tool that converts complex PDFs, images, and office files into structured Markdown, LaTeX, or JSON with advanced OCR and table recognition.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install mineru-precision-extract
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install mineru-precision-extract using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is mineru-precision-extract?
mineru-precision-extract is a professional-grade Openclaw Skills integration designed for deep document parsing. It leverages state-of-the-art models to transform static files—including academic papers, scanned PDFs, and multi-page reports—into editable, machine-readable formats. Whether you are building a RAG pipeline or digitizing archives, this skill provides the accuracy needed for production-level data engineering.
The tool offers a unique choice between the vlm model for maximum accuracy in complex layouts and the pipeline model for zero-hallucination reliability. By integrating directly with Openclaw Skills, developers can automate the conversion of 80+ languages and handle massive batch processing tasks through a unified CLI or agent interface.
mineru-precision-extract Use Cases
- Extracting complex tables and mathematical formulas from academic PDF papers into LaTeX.
- Converting scanned legal or historical documents into searchable Markdown using high-precision OCR.
- Automating batch document processing for large datasets in research and data engineering workflows.
- Crawling web pages to extract structured content while maintaining original layout integrity.
- Migrating legacy Word (DOCX) or PowerPoint (PPTX) files into structured documentation formats.
How mineru-precision-extract Works
- The user provides a local file path, a list of files, or a remote URL to the mineru-open-api.
- The skill identifies the document type and applies the selected model (vlm or pipeline) for layout analysis.
- OCR and specialized recognition engines extract text, recognize tables, and convert formulas into LaTeX strings.
- Extracted data is structured into the requested output formats such as Markdown, HTML, or DOCX.
- Images and assets are saved alongside the primary document, and the final result is delivered to the specified output directory.
mineru-precision-extract Setup
To get started with this component of Openclaw Skills, install the CLI via npm or Go:
npm install -g mineru-open-api
Or for macOS/Linux users:
go install github.com/opendatalab/MinerU-Ecosystem/cli/mineru-open-api@latest
After installation, authenticate by creating a token at the MinerU website and running:
mineru-open-api auth
mineru-precision-extract Data Schema & Taxonomy
The skill organizes extracted content into a structured hierarchy. When an output directory is specified, it creates the following taxonomy:
| File Type | Description |
|---|---|
*.md / *.json |
The primary structured text output containing the document content. |
images/ |
A subdirectory containing all figures and images extracted from the source. |
*.html / *.docx |
Optional alternative formats generated during the extraction process. |
metadata |
Internal taxonomy capturing page counts, language detection, and model logs. |
mineru-precision-extract Advanced Features
- Dual-Model Strategy: Choose vlm for intricate layouts or pipeline for verified, hallucination-free extraction.
- Massive Batch Processing: Support for processing hundreds of files concurrently via file lists or directory patterns.
- Intelligent Web Crawling: Convert live URLs directly into clean, structured Markdown for LLM training or archival.
- Global Language Support: Native recognition for over 80 languages across Latin, CJK, Arabic, and Devanagari scripts.
- Flexible Page Targeting: Use the --pages flag to extract specific sections from massive documents, saving time and API credits.
SKILL.md
Loading
Related Openclaw Skills
Issue Hunter
Issue Hunter is an advanced triage tool designed to systematically analyze, score, and select the best GitHub issues for developers to tackle.
Team Efficiency Daily Report
An automated workflow for capturing dashboard metrics and generating comprehensive daily team efficiency reports.
Claude Certified Architect (CCA) Learning Navigator
A comprehensive study guide and navigation hub designed to help developers pass the Anthropic Claude Certified Architect (CCA) certification.
WeChat Mini Program Development Skill
A comprehensive development toolkit for building, configuring, and optimizing high-performance WeChat Mini Programs.
ClawHub CLI
A command-line interface for the seamless discovery, installation, and publication of Openclaw Skills directly from clawhub.com.
First Aid Learning Assistant
An interactive educational agent designed to teach first aid principles, CPR techniques, and emergency preparedness through simulated drills and expert Q&A.
METADATA
Featured*
