A sophisticated multimodal tool designed to extract text and generate detailed descriptions from images using advanced AI models.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install image-reader
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 image-reader 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 Image Reader?
Image Reader is a specialized image recognition and understanding tool that integrates seamlessly into the Openclaw Skills ecosystem. By leveraging powerful multimodal models such as doubao-seed-2.0-pro or Kimi k2.5, this skill allows users to bridge the gap between visual media and actionable text. It is specifically engineered to handle diverse visual inputs, ranging from technical screenshots and UI designs to handwritten documents and complex scene photos.
This tool acts as a visual interface for AI agents, enabling them to perceive and interpret visual data as accurately as they process text. Whether you need to digitize a menu or understand the layout of a webpage, Image Reader provides the necessary technical infrastructure to perform high-fidelity OCR and scene analysis within your automated workflows.
Image Reader Use Cases
- Extracting code or error messages from screenshots for rapid debugging and technical support.
- Converting scanned documents, posters, and menus into editable text formats via high-accuracy OCR.
- Generating detailed accessibility descriptions or alt-text for images and illustrations.
- Analyzing UI/UX screenshots to identify components and visual styles during development reviews.
How Image Reader Works
- The user provides an image file path or a visual screenshot to the AI agent.
- The system determines the appropriate analysis strategy based on the user's specific request or the image type.
- The image is converted into a base64-encoded string for secure processing.
- A request is dispatched to an OpenAI-compatible multimodal API endpoint.
- The AI model processes the visual data and returns a structured response containing extracted text or visual descriptions.
- The final output is delivered back to the user or the next step in the Openclaw Skills workflow.
Image Reader Setup
To get started with Image Reader, ensure you have Python 3 and the necessary dependencies installed:
pip install "openai>=1.0.0" "pyyaml>=6.0"
Next, configure your API credentials in the config.yaml file, ensuring the endpoint is set to the provided multimodal service:
# Example execution for OCR
python image_reader.py /path/to/image.png -p "Extract all text from the image"
Image Reader Data Schema & Taxonomy
The Image Reader skill organizes data through a streamlined pipeline that converts visual inputs into structured text. The taxonomy of processed data is summarized below:
| Data Type | Description | Format |
|---|---|---|
| Input | Local image files or screenshots | PNG, JPG, JPEG |
| Intermediate | Encoded visual data | Base64 String |
| Output (OCR) | Plain text extracted from the image | UTF-8 Text |
| Output (Vision) | Descriptive analysis of objects and styles | Markdown String |
Image Reader Advanced Features
- Context-Aware Analysis: Automatically switches between OCR and Scene Description based on the detected density of text versus visual elements.
- Multimodal Model Support: Compatible with a variety of high-end vision models including Doubao and Kimi for flexible performance.
- Seamless Integration: Designed to be invoked via simple natural language commands within the Openclaw Skills environment, making visual intelligence accessible to non-technical users.
- Custom Prompting: Supports specialized instructions to focus the AI on specific areas of an image, such as reading only the header or identifying specific colors.
SKILL.md
Loading
Related Openclaw Skills
DNS Robot
A comprehensive toolkit for performing DNS queries, SSL audits, and email security checks without requiring an API key.
Live Shopping Notes
A specialized AI skill for capturing, tagging, and summarizing product offers and details during live commerce events.
Compliance Monitor
An automated regulatory tracking and deadline management tool designed for accounting and finance professionals.
Market News Analyst
An advanced AI skill that automates the collection and quantitative analysis of recent market-moving events affecting global equities and commodities.
Feishu Multi-Account Management
A powerful configuration skill that enables OpenClaw to route specific Feishu bot accounts to dedicated AI sub-agents.
Multi-Topic Deep Research
An intelligent tool that launches parallel agents to conduct deep, multi-topic research and generate structured knowledge bases.
Featured*
