← Back to Product Feed

GitHub Open Source ysr666/dsh-vision-router

Eyes for text-only DeepSeek Harness agents: built-in free vision chain (no key) + pixel-level vision tools (Q&A, grounding, crop, pixel diff, colors, OCR, SVG trace, cutout, screenshots). One-command install, no Python, image turns work like ordinary tool-calling turns.

990
Traction Score
44
Forks
Aug 13, 2026
Launch Date
View Origin Link

Product Positioning & Context

Eyes for text-only DeepSeek Harness agents: built-in free vision chain (no key) + pixel-level vision tools (Q&A, grounding, crop, pixel diff, colors, OCR, SVG trace, cutout, screenshots). One-command install, no Python, image turns work like ordinary tool-calling turns.
deepseek-harness dsh dsh-plugin multimodal vision

Related Ecosystem & Alternatives

Discover adjacent products, open-source repositories, and developer tools sharing similar technical architecture.

Deep-Dive FAQs

What is ysr666/dsh-vision-router?
ysr666/dsh-vision-router is a digital product or tool described as: Eyes for text-only DeepSeek Harness agents: built-in free vision chain (no key) + pixel-level vision tools (Q&A, grounding, crop, pixel diff, color...
Where did ysr666/dsh-vision-router originate?
Data for ysr666/dsh-vision-router was aggregated directly from the GitHub Open Source community ecosystem, representing raw developer and early-adopter sentiment.
When was ysr666/dsh-vision-router publicly launched?
The initial public indexing or launch date for ysr666/dsh-vision-router within our tracked developer communities was recorded on August 13, 2026.
How popular is ysr666/dsh-vision-router?
ysr666/dsh-vision-router has achieved measurable traction, logging over 990 traction score and facilitating 44 recorded discussions or engagements.
Which technical categories define ysr666/dsh-vision-router?
Based on metadata extraction, ysr666/dsh-vision-router is categorized under topics such as: deepseek-harness, dsh, dsh-plugin, multimodal.
Are there active development issues for ysr666/dsh-vision-router?
Yes, we are currently tracking open architectural debates and bug reports for this project on GitHub. There are currently 1 active high-priority issues logged recently.
What are some commercial alternatives to ysr666/dsh-vision-router?
Our semantic intelligence engine identifies potential commercial alternatives in the SaaS space, such as Arcjet, which offers overlapping value propositions.
How does the creator describe ysr666/dsh-vision-router?
The original author or development team describes the product as follows: "Eyes for text-only DeepSeek Harness agents: built-in free vision chain (no key) + pixel-level vision tools (Q&A, grounding, crop, pixel diff, colors, OCR, SVG trace, cutout, screenshots). One-comma..."

Active Developer Issues (GitHub)

open 从剪贴板粘贴 QQ/微信截图报「仅支持 PNG、JPG、WebP、GIF 格式的图片」
Logged: Aug 17, 2026

Community Voice & Feedback

xing666173 • Aug 19, 2026
https://github.com/xing666173/dsh-vision-hub
我的仓库也包含这个功能,欢迎共建
ysr666 • Aug 18, 2026
跟踪更新(2026-08-18):上游 Discussion deepseek-ai/deepseek-harness#2684 目前仍无官方回复;同时重新核对了官方当前 `master`,问题链路仍未变化:

- `packages/client/ui-conversation/src/client/service.ts` 的 `createDraftImages()` 仍直接用浏览器声明的 `file.type` 调 `imageMediaType()`,只接受 PNG/JPEG/WebP/GIF;
- `packages/client/ui-conversation/src/client/skeleton/InputBar.tsx` 的 `intakeImages()` 仍按 `file.type` 做整批预检,批内出现 `image/bmp` 就整批拒绝。

因此 Firefox/Windows 剪贴板把 DIB 暴露成 `image/bmp` 时,图片仍然会在进入附件栏之前被官方 composer 拒绝。这个阶段 Vision Router 还拿不到图片/附件,插件侧没有稳定、受支持的 API 可以做 MIME 归一化;强行做 DOM capture/paste monkey-patch 会和官方输入框生命周期、IME、拖拽/粘贴逻辑强耦合,回归风险明显高于收益。

结论:#138 继续保持打开并跟踪上游;暂不加入脆弱的 DOM 级 hack。最合理的修复仍应在官方 composer intake:按字节嗅探/把 BMP 转 PNG,并对同一剪贴板里的 DIB + FileDrop 做去重。
ysr666 • Aug 17, 2026
补充跟进:你后面反馈的 **Firefox 识图后持续 POST `/api/settings.mutate`、CPU 占用持续升高** 和本 issue 原本的剪贴板 BMP/MIME 拦截不是同一个问题。

这条已单独拆到 #155,并在 PR #156 修复合入 `main`:隐藏的 onboarding / 引导状态持久化现在会按原始 user settings 做同值去重,同一写入即使 Host 拒绝或 snapshot 反复刷新,也不会再被 subscribe 链路无限重发;真正的状态变化仍正常写入。完整测试已通过。

#138 原本的 QQ/微信剪贴板格式问题仍继续作为 deepseek-harness 上游问题跟踪。
ptxbc • Aug 17, 2026
firefox找了一会儿原因,是这个插件导致的,之前从桌面拉了一张图片识别,就一直cpu温度很高,网页一直post http://127.0.0.1:3080/api/settings.mutate ,清理127.0.0.1的缓存也不行,换个浏览器,没识别过图片就没问题。
ysr666 • Aug 17, 2026
感谢这份非常扎实的定位报告,根因分析我们完全认同 ✅

结论与你在「归属说明」中的判断一致:报错文案与校验逻辑都在 deepseek-harness 上游——

- 客户端 `packages/client/ui-conversation`:`imageMediaType()` 只精确匹配四个 MIME,`intakeImages()` 按整批拒绝;
- 服务端 `packages/attachment/attachment-local`:sharp 嗅探字节,声明与实际不符抛 `IMAGE_TYPE_MISMATCH`。

粘贴的图片在进入附件栏之前就被上游 UI 拦截,vision-router 插件层拿不到任何事件或数据,无法在插件侧兜底或绕过,因此本仓库无法单独修复,只能跟踪。

处理方式:

1. 你已在官方仓库开了 Discussion(https://github.com/deepseek-ai/deepseek-harness/discussions/2684 ),修复应在上游进行;你提出的按魔数嗅探字节、`image/bmp` 经 `createImageBitmap` → `canvas.toBlob("image/png")` 转码、粘贴批次去重等建议都很有价值,可一并提供给上游。
2. 本 issue 保持开放,作为该问题对 vision-router paste-and-go 流程影响的跟踪条目;上游修复发布后我们会回来回归验证并关闭。

给遇到此问题的用户的临时规避:

- Windows + Firefox 下不要直接 Ctrl+V 粘贴截图,先把截图保存为 PNG 文件(QQ/微信/系统截图都支持另存),再用附件按钮或拖拽上传;
- Chrome/Edge 下粘贴可用,但同图可能重复进入附件栏,提交前删掉重复项即可。

再次感谢对上游源码逐行定位的努力 👍

Discovery Source

GitHub Open Source GitHub Open Source

Aggregated via automated community intelligence tracking.

Tech Stack Dependencies

No direct open-source NPM package mentions detected in the product documentation.

Media Tractions & Mentions

No mainstream media stories specifically mentioning this product name have been intercepted yet.

Deep Research & Science

No direct peer-reviewed scientific literature matched with this product's architecture.