An AI-powered validation engine designed to review, OCR, and cross-validate diverse Korean business documents for compliance and accuracy.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install kr-document-reviewer
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 kr-document-reviewer 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 Korean Business Document Reviewer?
The Korean Business Document Reviewer is a specialized tool within the Openclaw Skills ecosystem designed to handle the complexities of Korean administrative and financial paperwork. It processes documents such as tax invoices, contracts, bank statements, and business registrations by converting them into machine-readable data and performing rigorous compliance checks. This skill is particularly valuable for businesses needing to ensure that financial figures, entity details, and dates remain consistent across multiple files in a transaction chain.
By leveraging advanced AI models, the skill can operate in two modes: a direct single-session mode for quick checks and a cost-optimized sub-agent mode for bulk processing. This flexibility makes Openclaw Skills an essential part of any modern financial operations stack, reducing human error in manual document review while significantly lowering processing costs through distributed agent architectures.
Korean Business Document Reviewer Use Cases
- Validating tax invoices against bank transfer confirmations to ensure payment accuracy.
- Cross-checking business registration certificates with contracts and estimates to verify vendor legitimacy.
- Auditing subsidy applications and result reports for consistency in project expense reporting.
- Bulk processing of transaction statements and delivery reports using parallel AI sub-agents.
How Korean Business Document Reviewer Works
- Pre-processing: The skill converts PDF files into high-resolution PNG images using pdftoppm to ensure optimal OCR performance.
- Document Identification: It analyzes the visual and textual content to identify the document type, such as a contract or a tax invoice.
- Parallel OCR: For large batches, it spawns sub-agents to extract data from multiple documents simultaneously, significantly reducing execution time.
- Cross-Validation: The skill compares extracted data (e.g., business numbers, amounts, dates) across the entire document set to detect discrepancies.
- Reporting: It generates a detailed JSON report and a human-readable summary highlighting critical issues and required actions.
Korean Business Document Reviewer Setup
To get started with this skill in your Openclaw Skills environment, ensure you have the necessary system utilities installed for PDF processing.
# Install poppler-utils for PDF-to-image conversion
sudo apt-get update
sudo apt-get install poppler-utils
Ensure that the references/ directory containing the document-specific validation templates and the schema/ directory for JSON output are present in your local workspace.
Korean Business Document Reviewer Data Schema & Taxonomy
The skill produces a structured JSON output based on a predefined schema to facilitate integration with other Openclaw Skills.
| Component | Description |
|---|---|
| meta | Includes reviewId, reviewDate, and a high-level transaction summary. |
| documents | List of individual document results including extracted data and pass/fail status for specific fields. |
| crossValidation | Result of consistency checks between documents, such as matching the contract amount to the invoice. |
| summary | Final assessment including total document count, warning/fail counts, and specific action items. |
Korean Business Document Reviewer Advanced Features
- Sub-agent Task Delegation: Utilizes sessions_spawn to distribute OCR tasks to Claude Sonnet sub-agents, reducing costs by approximately 94% compared to main-session processing.
- Korean OCR Optimization: Includes logic to correct common OCR misreadings of Korean names, company titles, and complex addresses.
- Flexible Output Formatting: Saves comprehensive JSON logs for audit trails while providing formatted summaries suitable for Discord or team chat notifications.
- Multi-page PDF Support: Handles multi-page documents by processing each page and aggregating the extracted data logically.
SKILL.md
Loading
Related Openclaw Skills
NEAR QR Code Generator & Reader
A Python-powered utility for generating and decoding NEAR Protocol QR codes for addresses and payment requests.
iOS Keyboard Extension Technical Guide
A technical blueprint detailing the constraints and workarounds for building iOS custom keyboard extensions, specifically for audio and AI features.
Home Assistant Integration Patterns
A comprehensive architectural guide for building robust Home Assistant custom integrations and HACS components.
Scrappa MCP
A comprehensive Model Context Protocol server providing AI agents with instant access to over 80 web scraping and data retrieval tools.
ASCII Art Generator
Generate intricate ASCII art, decorative text banners, and conceptual diagrams using characters, symbols, and whitespace.
Download Organizer
A command-line utility that automatically identifies and organizes downloaded files into category-specific folders based on file type.
METADATA
Featured*
