DeepSeek Harness Plugin Hub
← Plugins

dsh-vision-assist

Vision Assist

DeepSeek Harness vision assistant plugin: equips models without vision capabilities with a switchable multimodal recognition model. Images in the input field are automatically saved to disk and rewritten as text prompts; the main model can view images by calling the vision_recognize tool. The recogn

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

npx -y @deepseek-ai/dsh plugin --profile web add github:wjling/dsh-vision-assist#072d8f77f12c5037a4f8e81adb9346f6b4e4a1d8

Description

DeepSeek Harness vision assistant plugin: equips models without vision capabilities with a switchable multimodal recognition model. Images in the input field are automatically saved to disk and rewritten as text prompts; the main model can view images by calling the vision_recognize tool. The recognition model can be hot-switched in the vision-assist namespace under settings.

Compatibility and provenance

Vision Assist is published as dsh-vision-assist and currently resolves to version 1.0.0. 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

1.0.0stable