loudMore/dsh-drop-to-path ↗★ 6
@dsh-external/dsh-drop-to-path
支持将图片、PDF、Office 文档、压缩包、音视频拖放或粘贴至 DSH 输入框，并转换为工作区文件路径而非模型附件，以便纯文本模型通过视觉工具读取。
AI 分析
核心用途是将拖放文件转换为工作区路径。适合配合 dsh-vision-toolkit 使用，让纯文本模型能通过工具间接读取各类媒体与文档。
安裝
$
npx -p @deepseek-ai/dsh dsh plugin --profile web add github:loudMore/dsh-drop-to-path說明文件
閱讀完整 README ↗Usage
- Paste or drop images as usual — the attachment cards behave natively (preview / remove);
- Drag or paste any other file (pdf / office / zip / video / audio) — a square chip with a format icon appears in the attachment rail next to the image thumbnails (full name on hover, ✕ to remove);
- Type your message (optional) and hit send — no attachments are sent; images and files are uploaded and their workspace paths are appended to the message automatically;
- The agent (paired with dsh-vision-toolkit) reads images with
vision_glance/vision_pixel_diffand friends, and handles documents with PDF / spreadsheet tools.
Files are stored under /.drops/; clean them up whenever you like.