dsh-eyes
On-demand vision for text-only DeepSeek models: upload images, and the model calls a view_image tool backed by any OpenAI-compatible vision endpoint (Qwen/DashScope by default).
为纯文本 DeepSeek 模型提供按需视觉：上传图片后，模型通过 view_image 工具调用任意 OpenAI 兼容视觉端点（默认 Qwen/DashScope）。
How to install
dsh plugin add dsh-eyes About
dsh-eyes **中文** · English 给 DeepSeek Harness 里的**纯文本大模型**（如 DeepSeek）装上「随时可用的眼睛」：图片粘贴/附件后留在后台，模型自己在需要时调用 view_image 工具去看图（底层走**任意 OpenAI 兼容视觉接口**，默认百炼 Qwen），就像模型**原生具备多模态**一样。 快速体验：上传图片，自然丝滑 不需要开关、不需要手动调用工具——**直接 Ctrl+V 粘贴一张截图**，像发普通消息一样提问即可。DeepSeek 会在思考里自然地决定「我要看这张图」，自己调用 view_image，然后直接给出结构化回答，一气呵成： 上图实拍：粘贴一张 MSN 截图，问「介绍这个页面的布局」。注意中间的 Think → Tool call · view_image → Think 过程——识别不是被强行塞进第一步，而是模型在需要时自己决定看图；视觉提取与最终回答无缝衔接，就像 DeepSeek **原生具备多模态**一样丝滑。 解决的问题 DeepSeek 是纯文本模型，Harness 默认不允许给「当前模型不支持图片」的会话发送带图消息。本插件： 让带图消息**能通过发送准入**并被持久化保存； 在把消息交给主模型**之前**把图片剥离成一句引用说明； 注册 view_image 工具，主模型**随时**调用它看…
Recommendation signals
Meta
- License
- MIT
- Language
- JavaScript
- GitHub stars
- 1
- mo. downloads
- 266
- Last push
- 2026-08-16
- Created
- 2026-08-16
Links
Basic safety check
- Findings
- None
- Sources
- curated:awesome-dsh-plugin.com, curated:awesome-dsh-plugin/awesome-dsh-plugin
- Topics
- dashscope, deepseek-harness, dsh-plugin, multimodal, ocr, openai-compatible, qwen, vision
Related plugins
modlens
liustack/modlens
Vision bridge for text-only models: paste an image, get structured JSON evidence (OCR, layout, semantics).
dsh-web
zhu1090093659/dsh-web-ui/tree/main/packages/dsh-tool-describe-image
A `describe_image` vision tool for text-only models: images (local path, URL, attachment) go to a configurable OpenAI-compatible vision endpoint and only the returned text enters the session.
dsh-vision-router
ysr666/dsh-vision-router
Free vision for text-only agents: built-in keyless vision chain plus pixel tools (Q&A, grounding, crop, pixel diff, colors, OCR, SVG trace, cutout, screenshots); paste an image to use it.
dsh-vision-toolkit
Anionex/dsh-vision-toolkit
Vision for text-only models: paste an image and the model switches to a Vision Toolkit variant for image Q&A, multi-image comparison, long-screenshot OCR, screenshot-to-UI reproduction, element grounding, and pixel diff. No API key by default — images are processed by the author-hosted free service, 100 per machine per day; configurable to your own provider.