Agent Vision Toolbox
loudMore/dsh-drop-to-path · skills
Drop or paste images, PDFs, office docs, zips, videos and audio into DeepSeek Harness — they reach your text-only model as workspace file paths.
★ 11GitHub stars
1Forks
2026-08-15Last updated
JavaScriptLanguage
MITLicense
Key features
- Drag images / PPT / TXT / ZIP into the composer → square file chips appear next to image thumbnails → hit send → the model receives workspace file paths.
- The image admission preflight (attachment-error in dsh-host-apiproxy) is bypassed: no image block is ever sent, the model only sees text.
- dsh-drag-and-dropdropping a file inserts its raw path into the composer, no file copying.
- images ≤30MB, other files ≤100MB
- The effectimages and files arrive at the text-only model as workspace paths, ready for vision / document tools.
Requirements
- This plugin depends on several undocumented internal interfaces of DSH (service names, method signatures, storage formats) and may break after a DSH upgrade.
- Read ADAPTING.md before upgrading — it contains the failure symptoms and the fix steps.
Install command
dsh plugin --profile web add github:loudMore/dsh-drop-to-path