DeepSeek Harness Plugin Hub
← Plugins

@xiaoyuink/dsh-image-vision

Image Vision

Image recognition plugin: Automatically detects whether the current model has vision capabilities. If so, it uses the current model with the plugin’s preset prompts for analysis; otherwise, it calls the vision model configured in the plugin. Text models can paste, upload, or drag images directly int

The plugin will be installed here. Keep web if you are unsure.

npx -y @deepseek-ai/dsh plugin --profile web add github:xiaoyuink/dsh-image-vision#a445f8437a9e1cdddaf4ded34b5d863bd1785b0d

Description

Image recognition plugin: Automatically detects whether the current model has vision capabilities. If so, it uses the current model with the plugin’s preset prompts for analysis; otherwise, it calls the vision model configured in the plugin. Text models can paste, upload, or drag images directly into the chat box. When sent, images are automatically saved to DSH attachment storage permanently, with thumbnails rendered in the message area. The model automatically calls the image_vision / ocr / ground / crop tool series for recognition and detailed analysis, while keeping the model selector simple.

Compatibility and provenance

Image Vision is published as @xiaoyuink/dsh-image-vision and currently resolves to version 2.9.2. The Hub verifies its manifest and preserves the exact installation source for reproducible installs.

DSH compatibility
*
Runtime surfaces
web
Release source
github
Registry updated

Versions

2.9.2stable
2.9.0stable
2.8.0stable
Show 3 more versionsCollapse versions
2.7.2stable
2.6.1stable
2.5.0stable