image-analysis

Perform OCR and object detection on local images by providing the file path and a descriptive prompt. The skill processes local inputs, normalizes data, and routes requests to configured vision models for results.

Yes, you can analyze network images and video files for scene understanding. The tool handles input normalization and web inputs, routing them to vision models to extract contextual scene information.

Yes, you must configure API keys and default vision models in scripts/config.json before running image analysis. This setup ensures proper model routing and successful processing of OCR and object detection tasks.

Yes, this image analysis tool supports both GLM-4V and Qwen-VL vision models. This dual-model support ensures you can leverage different visual understanding capabilities across various analysis scenarios.

The best way to extract text from images without pytesseract is using this dedicated skill. It provides a unified entry point for OCR tasks, directly calling configured vision models to bypass other Python image libraries.

Image analysis requires input normalization and model routing to handle diverse local and web inputs consistently. This process ensures accurate error reporting and directs requests to the correct configured vision model.