A professional AI skill for converting handwritten notes, letters, and mixed-content documents into editable digital text using the ZhiPu GLM-OCR API.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install glmocr-handwriting
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 glmocr-handwriting 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 GLM-OCR Handwriting Recognition?
The GLM-OCR Handwriting Recognition skill is a specialized tool within the Openclaw Skills ecosystem designed to bridge the gap between physical handwriting and digital data. By leveraging the ZhiPu GLM-OCR layout parsing API, it accurately interprets various handwriting styles, including cursive and print, across multiple languages like Chinese and English.
This skill is particularly effective for processing complex documents that contain both handwritten and printed text. It ensures high fidelity in transcription by focusing strictly on professional OCR endpoints, making it a reliable choice for developers and researchers who need to digitize physical archives or personal notes without manual data entry.
GLM-OCR Handwriting Recognition Use Cases
- Extracting text from handwritten personal letters, academic notes, or historical documents.
- Converting physical meeting minutes or brainstormed whiteboards into editable Markdown text.
- Processing official forms that contain a mix of printed labels and handwritten input.
- Digitizing handwritten mathematical formulas, labels, or marginal annotations in textbooks.
- Automating the ingestion of handwritten feedback or surveys into digital databases.
How GLM-OCR Handwriting Recognition Works
- The user provides a local file path or a remote URL pointing to an image or PDF containing handwriting.
- The skill validates the presence of the ZHIPU_API_KEY to ensure authorized access to the layout parsing engine.
- A Python-based CLI script sends the file data to the fixed ZhiPu GLM-OCR endpoint for processing.
- The API analyzes the layout and performs deep character recognition on the handwriting.
- The skill receives a structured JSON response containing the full extracted text and layout details.
- The final output is presented to the user in a clean Markdown format or saved to a specified local file.
GLM-OCR Handwriting Recognition Setup
To use this feature within Openclaw Skills, you must configure your ZhiPu API key. You can obtain your key from the ZhiPu AI Open Platform.
Environment Configuration
Set your API key in your global openclaw.json or as an environment variable:
# Add to your shell profile (e.g., ~/.zshrc or ~/.bashrc)
export ZHIPU_API_KEY="your_api_key_here"
Execution
Run the recognition script using Python:
# From a local file
python scripts/glm_ocr_cli.py --file /path/to/handwriting.png
# From a remote URL
python scripts/glm_ocr_cli.py --file-url "https://example.com/note.jpg"
GLM-OCR Handwriting Recognition Data Schema & Taxonomy
The skill returns a structured response that ensures data integrity and ease of integration.
| Field | Type | Description |
|---|---|---|
ok |
Boolean | Indicates if the OCR process was successful. |
text |
String | The primary extracted content formatted in Markdown. |
layout_details |
Array | Objects containing coordinate and structural data for identified elements. |
source |
String | The original path or URL of the processed file. |
error |
String | Detailed error message if ok is false. |
raw_result_included |
Boolean | Flag indicating if the full API response was attached for debugging. |
GLM-OCR Handwriting Recognition Advanced Features
- Multi-language Support: Seamlessly handles documents containing mixed Chinese and English handwriting.
- Layout Parsing: Goes beyond simple text extraction to understand the physical structure of the document.
- Debug Mode: Use the
--include-rawflag to inspect the upstream API response for fine-tuning results. - Flexible Output: Supports pretty-printing JSON results directly to the console or saving them to local storage for automated workflows.
- Security Hardening: Uses fixed official endpoints to prevent accidental API key exposure through URL redirection.
SKILL.md
Loading
Related Openclaw Skills
OpenClaw Panel Control
A comprehensive control interface for ESP32-S3 based 64x32 RGB LED panels over local network HTTP requests.
text-to-ppt
A high-performance AI skill that converts raw text into visually stunning, interactive HTML presentations using a sophisticated two-phase generation process.
Upload Skill for Payments
A comprehensive automation skill for creating, managing, and testing diverse payment structures and subscription models within a secure sandbox environment.
Zhipu AI Hot Trend Tracker
A specialized automation skill for monitoring and synthesizing public technical documentation, pricing models, and feature updates from Zhipu AI.
Image AI Kit
A comprehensive AI-powered toolkit for intelligent image enhancement, super-resolution, and automated OCR processing.
Video Resizer AI Agent Skill
A specialized Openclaw Skills variant for Sparki designed to automate video aspect ratio conversion and platform-specific formatting.
Featured*
