npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

dsh-vision-pro-bridge

v1.0.0

Published

Give text-only DeepSeek-V4-Pro real vision with zero new dependencies and DeepSeek-only routing: images are described by deepseek-v4-flash-vision-exp (your existing DEEPSEEK_API_KEY), then the text is handed to V4-Pro.

Readme

dsh-vision-pro-bridge

让纯文本的 DeepSeek-V4-Pro 真正“看见”图片 —— 不用切换模型、不用 Ollama、不用第二个 key。

解决的问题

deepseek-v4-pro(以及 deepseek-v4-flash)是纯文本模型:DeepSeek API 不接受图片字节。在 DeepSeek Harness 里,往 Pro 会话里贴图会报“不支持图片”。官方唯一的选择,是把整个会话切到(编码能力较弱的 Flash 档)deepseek-v4-flash-vision-exp。

这个插件做什么

它注册一条孪生路由 deepseek-vision-pro(模型 deepseek-v4-pro-vision)并声明支持图片。你贴图后,插件先用 deepseek-v4-flash-vision-exp 把图片转成文字(复用你已有的 DEEPSEEK_API_KEY),再把文字 + 本地路径交给 deepseek-v4-pro。DeepSeek 的对话路由永远收不到图片,V4-Pro 保留编码能力、同时“看得见”。

贴图 → [图片块被放行]
   → deepseek-v4-flash-vision-exp 转写为文字
   → 文字 + 本地路径 → deepseek-v4-pro 回答

为什么选这个(优势)

  • 零第三方依赖:只复用 Harness 自带的 @deepseek-ai/dsh-llm 与 @deepseek-ai/dsh-llm-deepseek。没有 openai、没有 sharp、没有 schemastery、没有 Ollama。
  • DeepSeek 专属、单一供应商:图片只发给 DeepSeek(deepseek-v4-flash-vision-exp),走同一个 DEEPSEEK_API_KEY。没有第二个 VLM 厂商、没有本地模型、没有匿名免费端点。
  • 官方图片管线、逐字节一致:继承官方 DeepSeekAdapter,图片规范化、Files API 上传、inline-base64 回退都与官方行为完全一致,不手写 HTTP。
  • 极简、可审计:单文件、约 250 行,易读、易验、易 fork。
  • 内容寻址落盘:图片存到 ~/.dsh/vision-pro-bridge/images/,稳定本地路径会传给模型,V4-Pro 可继续引用。
  • Pro 专精:只暴露一个模型 deepseek-v4-pro-vision —— 保留 V4-Pro 编码能力,外加视觉。

安装

dsh plugin --profile web add dsh-vision-pro-bridge
# 重启 dsh web

使用

  1. 模型选择器里选 DeepSeek-V4-Pro (视觉桥)。
  2. 粘贴/拖入截图,正常提问。

设为新会话默认(~/.dsh/settings.yaml):

agent-default-model:
  provider: deepseek-vision-pro
  model: deepseek-v4-pro-vision

配置(环境变量)

| 变量 | 默认 | 作用 | | --- | --- | --- | | DEEPSEEK_API_KEY | credentials 服务 | DeepSeek key(与官方路由同源) | | DEEPSEEK_BASE_URL | https://api.deepseek.com | 端点(与官方路由同源) | | DSH_VISION_PRO_BRIDGE_PROMPT | 内置英文转写提示词 | 覆盖转写提示词(例如改为中文 UI 分析) |

已知限制

  • 转写是有损的:布局/OCR 足够,精确像素坐标不如真视觉。
  • 转写模型固定为 deepseek-v4-flash-vision-exp。
  • 端点从 DEEPSEEK_BASE_URL 解析,不读设置页里 llm-deepseek 的自定义项。
  • 无降级链:转写失败时图片变成占位文本,对话继续。

备选

需要更完整功能(VLM 降级链、本地 Ollama、设置页、图片降采样)请看 dsh-vision-proxy。本插件是同一思路的极简、纯 DeepSeek、零依赖版本。

License

MIT