A high-performance OCR tool utilizing PaddleOCR to extract Chinese and English text from images and PDF documents.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install ocr-python
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 ocr-python 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 OCR Text Recognition?
The OCR Text Recognition skill is a specialized utility designed for developers and AI agents needing to bridge the gap between visual media and structured text. By leveraging PaddleOCR, it provides robust support for both Chinese and English characters, making it ideal for processing international documents and complex layouts. This skill is a core component of the Openclaw Skills ecosystem, enabling the seamless transformation of scanned PDFs, invoices, and photos into machine-readable data.
Whether you are building an automated bookkeeping agent or a document archiving system, this skill provides the underlying engine for accurate character recognition. It simplifies the integration of advanced computer vision capabilities into your development workflow without requiring deep expertise in machine learning models.
OCR Text Recognition Use Cases
- Scanned PDF processing: Extract text from multi-page non-searchable PDF documents by converting pages to images first.
- Document digitization: Identify and extract text from invoices, contracts, and official forms for further data processing.
- Image-to-text conversion: Rapidly retrieve text content from photographs, screenshots, or receipts for archival or analysis within Openclaw Skills pipelines.
How OCR Text Recognition Works
- Initialize the PaddleOCR engine with the desired language parameters (e.g., 'ch' for Chinese and English).
- Provide the file path of the target image or PDF to the prediction engine.
- The skill processes the file using the underlying prediction model to identify and segment text regions.
- Structured results are returned, containing the extracted strings and their corresponding confidence scores for programmatic validation.
OCR Text Recognition Setup
Install the necessary dependencies via pip to prepare your environment for Openclaw Skills integration:
pip3 install paddlepaddle paddleocr
Ensure that your system has the appropriate drivers if you intend to use GPU acceleration for larger batches of documents.
OCR Text Recognition Data Schema & Taxonomy
The skill returns a structured JSON object containing the results of the recognition process, ensuring compatibility with other tools in the Openclaw Skills ecosystem.
| Field | Type | Description |
|---|---|---|
rec_texts |
Array | A list of all recognized text strings found in the document. |
rec_scores |
Array | Numerical confidence values for each corresponding recognized text string. |
OCR Text Recognition Advanced Features
- Multi-language support: Specifically optimized for high-accuracy recognition of both Chinese and English character sets.
- PDF workflow integration: Designed to work alongside PyMuPDF to handle complex, multi-page document scanning and image extraction.
- Confidence filtering: Leverage the rec_scores to programmatically filter or flag low-quality extractions in Openclaw Skills automation scripts.
- Extensible scripting: Includes a library of common scripts in the scripts/ directory for rapid deployment in production environments.
SKILL.md
Loading
Related Openclaw Skills
Brave Search
A high-performance tool for headless web search and markdown content extraction using the Brave Search API.
RNWY Agent Registry — A2A Card Lookup
A verification and discovery engine for AI agents that provides live A2A agent cards and trust scores for over 40,000 entities.
beads
A distributed, git-backed graph issue tracker specifically designed for AI agents to manage complex task dependencies and project state.
ContextClaw
ContextClaw is a comprehensive session management and cleanup utility designed to optimize storage and analyze token consumption for AI agents.
Aperture
A high-performance reverse proxy that implements the L402 protocol to enable Lightning Network payment-gated API access.
Commerce Toolkit
An end-to-end toolkit for agentic commerce using the Lightning Network and L402 protocol to enable autonomous micropayments.
Featured*
