A high-performance OCR skill for extracting text from images, screenshots, and scanned documents in over 100 languages using PaddleOCR.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install smar
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 smar 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 Smart OCR?
Smart OCR is a robust text-extraction tool designed for the Openclaw Skills framework. It leverages the state-of-the-art PaddleOCR engine to identify and convert visual text into machine-readable data. Whether you are dealing with complex business cards, handwritten notes, or multi-page scanned PDFs, this skill provides the precision needed for modern AI workflows.
By integrating this tool into your Openclaw Skills setup, you gain access to features like angle classification, layout reconstruction, and multilingual support. It is optimized for both CPU and GPU environments, ensuring fast processing times for single images or large batches of documents.
Smart OCR Use Cases
- Digitizing text from screenshots or business card photos.
- Extracting content from multi-page scanned PDF documents.
- Automating receipt scanning and financial data entry.
- Analyzing text from multilingual documents containing languages like Chinese, Japanese, or Arabic.
- Preprocessing visual data for search indexing within Openclaw Skills.
How Smart OCR Works
- The user provides an image file, PDF, or URL to the Openclaw Skills agent.
- The skill optionally detects the document language or uses a pre-specified language code.
- PaddleOCR processes the input to identify text blocks, bounding boxes, and confidence scores.
- Advanced layout reconstruction logic organizes the raw text into logical lines and paragraphs.
- The agent returns the structured text data or a summary to the user.
Smart OCR Setup
To get started with this skill in your Openclaw Skills environment, install the necessary dependencies:
# Install the OCR engine (CPU version)
pip install paddlepaddle paddleocr
# For GPU acceleration (optional)
pip install paddlepaddle-gpu
# Install additional document processing tools
pip install pdf2image Pillow
Smart OCR Data Schema & Taxonomy
The Smart OCR skill organizes its output into structured data, tracking spatial coordinates and text accuracy.
| Attribute | Description |
|---|---|
| text | The actual string extracted from the image. |
| confidence | A decimal score (0.0 - 1.0) representing extraction accuracy. |
| bbox | Bounding box coordinates (left, top, right, bottom). |
| raw_box | The four-point polygon vertices of the text area. |
| language | The detected or specified language code used for OCR. |
Smart OCR Advanced Features
- Angle Classification: Automatically detects and corrects rotated text for higher accuracy.
- Layout Reconstruction: Groups disparate text blocks into logical, readable paragraphs based on vertical and horizontal proximity.
- Image Preprocessing: Built-in support for contrast enhancement and sharpening using Pillow to handle low-quality scans.
- Batch Processing: Parallel execution support for high-throughput document analysis within Openclaw Skills.
- Multilingual Fusion: Ability to merge results from different language models for complex documents.
SKILL.md
Loading
Related Openclaw Skills
AI Image Generation (gen-image)
A high-performance skill for generating AI images using industry-leading models like FLUX and Kolors through the SiliconFlow API.
Intercom Conversations Read
A specialized tool for listing, finding, and searching Intercom customer conversations via a structured JSON interface.
Apify Web Scraping & Data Extraction
A powerful tool for discovering and executing Apify Actors to scrape structured data from social media and websites.
DrawThings Image Generation
A high-performance local image generation skill that interfaces with DrawThings to provide Stable Diffusion capabilities via an API on macOS.
Movie Butler
An AI-powered cinema assistant that bridges TMDB data with Emby and Plex servers for intelligent media management and context-aware recommendations.
Snail Mail (Operator Inbox)
A dedicated slow-channel notification system that queues non-urgent agent updates for structured human review.
Featured*
