DSH Plugins Marketplace

DSH Plugins

Plugins

/

Tools & Capabilities

/

dsh-window-locator

d

dsh-window-locator

Identified

Exposes Windows display geometry and the foreground window rect in physical pixels.

dsh-window-locator

Windows 显示器几何与前台窗口查询(物理像素),做成 DeepSeek Harness 工具。

社区插件,与 DeepSeek 官方无关。
概念来自 canyueY/window-locator。
English: README.en.md

为什么需要它

截图工具和点击工具最怕坐标系不一致:Qt 的 logical 坐标、浏览器的 CSS 像素、Windows 的物理像素,在 125% / 150% 缩放下会差出一大截。这个插件统一返回 物理像素:

  • 进程内先 SetThreadDpiAwarenessContext(PMv2),再枚举显示器与前台窗口;
  • 不经过任何 Qt 逻辑坐标换算;
  • 多显示器时给出各自的 bounds 与 work 区域,主屏单独标出。

典型用法:让 Agent 先 desktop_snapshot,知道把鼠标点到哪、截图裁哪、当前前台是不是自己。

特性

  • 多显示器:device / bounds / work / is_primary;
  • 前台窗口:hwnd / title / rect / pid / 是否最大化 / is_self(是不是 DSH 自己);
  • 一次拿全:desktop_snapshot = 显示器列表 + 主屏 + 前台窗口;
  • 仅 Windows:底层通过 PowerShell 调 Win32 API。

前提

  • Windows 10 / 11;
  • powershell.exe 可用(系统自带即可);
  • Node ≥ 20(只影响开发,插件运行时由 DSH 提供)。

安装

dsh plugin --profile web add github:DSH-Raising-Plan/dsh-window-locator#plugin/dsh

工具

工具参数作用
list_monitors无列出全部显示器(物理像素 bounds / work / 主屏标记)
foreground_window无前台窗口标题、矩形、pid、最大化状态
desktop_snapshot无一次返回显示器列表 + 主屏 + 前台窗口

输出示例(节选):

{
  "monitors": [
    {
      "device": "\\\\.\\DISPLAY2",
      "bounds": { "x": 0, "y": 0, "width": 2560, "height": 1440 },
      "work": { "x": 0, "y": 0, "width": 2560, "height": 1392 },
      "is_primary": true
    }
  ],
  "primary": { "device": "\\\\.\\DISPLAY2", "is_primary": true },
  "foreground": {
    "hwnd": 67372,
    "title": "…",
    "rect": { "x": 109, "y": 139, "width": 1523, "height": 893 },
    "pid": 12345
  }
}

使用示例

desktop_snapshot  → 看有几块屏、主屏多大、当前前台窗口是谁
foreground_window → 截图前确认前台不是自己,避免把 Agent 的窗口截进去

限制

  • 仅 Windows;其它平台直接报错;
  • 每次调用会启动一次 PowerShell,首次约几十到一百多毫秒;
  • 返回的是物理像素,不是 DPI 无关的逻辑像素;如果你的下游按逻辑坐标算,需要自己除 scale;
  • 某些全屏独占 / 受保护内容 / 提权窗口可能拿不到准确矩形或标题;
  • 不提供“按标题找窗口”的枚举接口,只查前台。

开发

npm test
npm run check

需要 @deepseek-ai/dsh-tools(或设置 DSH_TOOLS_PATH)。

许可

AGPL-3.0-or-later

Comments

Loading…

Similar plugins

dsh-computer-use-windows

by Altairpaca

Windows Computer Use for DeepSeek Harness (DSH): window-bound screenshot/OCR/click with verification loop, pure-OCR mode, pluggable vision models.

Tools & CapabilitiesManifest valid

★ 3

↓ 113/wk

MIT

PowerShell

Sep 3, 2026

dsh plugin --profile web add dsh-computer-use-windows

by 7gzgn

Windows GUI automation reliability layer: probes whether a window is minimized, foreground or DPI-scaled, then clicks and types with the coordinate space and activation order handled for you. Zero dep

Tools & CapabilitiesManifest valid

★ 0

MIT

JavaScript

Oct 17, 2026

dsh plugin --profile web add dsh-gui-guard

by zzy6-a

Desktop computer-use for DSH: captures the Windows screen into the agent vision channel and drives mouse/keyboard through a Codex-style overlay with Esc abort; auto-detects Windows native or WSL hosts

Tools & CapabilitiesManifest valid

★ 4

↓ 294/wk

MIT

PowerShell

Oct 27, 2026

dsh plugin --profile web add dsh-vision

Windows-native computer use for DeepSeek Harness: two tools drive and read any desktop app. One batch call runs find, click, type, key, read, wait and screenshot steps in a single engine pass, so a ta

Tools & CapabilitiesManifest valid

★ 0

dsh plugin --profile web add dsh-win-computer-use

by MitsukiJoe

Display companion for dsh-vision-router: square picture button on each original model row

Tools & CapabilitiesManifest valid

★ 0

MIT

JavaScript

Aug 30, 2026

dsh plugin --profile web add dsh-vision-router-inline

by cbg33695

Let a text-only model see the screen: capture desktop/window/region, vision transcription, a few minutes of rolling screen memory, exact local pixel diff, and vision self-calibration. Windows only. Ex

Tools & CapabilitiesManifest valid

★ 0

MIT

JavaScript

Sep 30, 2026

dsh plugin --profile web add dsh-screen-reader