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

@fuxi-ailabs/dsh-insight

v1.1.0

Published

DSH integration for Insight: Session import, scoring commands, and a floating Session launcher

Readme

@fuxi-ailabs/dsh-insight

dsh-insight 是 DSH 的 Insight 宿主插件。

当前提供以下能力:

  • /insight:导出当前 DSH 会话,通过 Insight API 导入,并可打开会话页。
  • /score:读取、重算或语义评审会话评分卡,并做导入覆盖度对账。
  • 会话右下角浮动入口:点击后刷新导入当前会话,并打开 Insight 会话页。
  • 状态接口:GET /api/dsh-insight/status?sessionId=<sid>。
  • 打开接口:POST /api/dsh-insight/open。
  • 浮窗打分接口:POST /api/dsh-insight/score(默认 /score 同一条链路,不调用 LLM)。
  • 评分卡数据口:GET /api/dsh-score/scorecard,供后续评分卡 UX 接入。

安装

pnpm --filter @fuxi-ailabs/dsh-insight pack

dsh plugin --profile <profile> add \
  ./fuxi-ailabs-dsh-runtime-bundle-1.1.0.tgz \
  ./fuxi-ailabs-dsh-insight-bundle-1.1.0.tgz

dsh-insight 依赖 environmentHealth 服务。它把 Insight 安装请求交给 @fuxi-ailabs/dsh-env-health 的 Runtime Provisioner,由共享服务对同一版本的并发请求做单飞和跨进程锁。 首次准备通过 npm 下载 [email protected] 到 ~/.dsh-insight/runtime/;下载不阻塞 DSH 启动,也只会执行一次。首次点击浮窗或执行 /insight、/score 时才启动 Insight 服务。

运行时需要本机可找到 npm。找不到时共享 Runtime Provisioner 会返回错误,不影响 DSH 启动。

/insight

/insight                                   当前会话(含子代理)
/insight <sid> [<sid> ...]                 指定会话
/insight --out <dir>                       覆盖本次导出目录
/insight --no-subagents                    本次不带子代理
/insight --no-open                         导入后不开浏览器
/insight --base <url>                      指定 Insight 基址
/insight --force                           Insight 未运行时仍导出
/insight --status                          只探活

流程:

探活 Insight
→ 导出 canonical DSH JSONL
→ POST /api/ingest/import-file
→ 打开 {base}/session/<sid>?framework=dsh

/score

/score                                     读卡;没有缓存则现算确定性层
/score --sid <sid>                         指定会话
/score --recalc                            强制重算确定性层
/score --semantic                          语义评审(唯一 LLM 入口)
/score --probe-provider                    探测语义评审 provider
/score --confirm D2.destructive=fine       人工确认疑似项
/score --bench <0-100>                     填 CANN-Bench 总分
/score --open                              打开 Insight 页面
/score --json                              保存整卡 JSON
/score --no-cover                          跳过覆盖度对账
/score --base <url>                        指定 Insight 基址
/score --status                            只探活

/score 的确定性读卡与重算不调用 LLM;只有显式使用 --semantic 才会请求 Insight 的语义评审接口。

配置

- id: dsh-insight
  name: '@fuxi-ailabs/dsh-insight'
  config:
    # /insight
    outputRoot: "~/.dsh-insight/dsh-export"
    insightBin: context-insight
    autoInstall: true
    autoStart: true
    runtime:
      packageName: context-insight
      version: 0.2.41
      installRoot: "~/.dsh-insight/runtime"
      npmBin: npm
      registry: "https://registry.npmjs.org/"
      installTimeoutMs: 1200000
      startTimeoutMs: 180000
    insightBase: "http://127.0.0.1:21025"
    portFile: null
    includeDescendants: true
    openBrowser: true
    requireInsightUp: true
    source: dsh-native
    probeTimeoutMs: 1500
    importTimeoutMs: 120000

    # /score(独立命名空间,避免 outputRoot 冲突)
    score:
      outputRoot: "~/.dsh-insight/dsh-score"
      framework: dsh
      includeDescendants: true
      coverToleranceMs: 2000
      requestTimeoutMs: 60000
      semanticTimeoutMs: 300000
      openBrowser: false
      openTarget: scoring
      provider: null
      borrowDshProvider: true
      dshProvider:
        apiKeyRef: DEEPSEEK_API_KEY
        baseUrl: "https://api.deepseek.com"
        model: "deepseek-flash"

浮窗交互

会话页右上角显示可拖拽的 Insight 浮标,默认位置和拖动结果保存在浏览器本地。

  • Insight 未启动:显示 ⊘ Insight,点击提示“Insight 实例没有启动”。
  • Insight 已启动、会话未导入:显示普通 Insight 按钮,点击刷新导入并打开会话页。
  • 会话已导入:Insight 收缩为图标,右侧出现「打分」按钮。
  • 点击「打分」:执行默认 /score 链路,不调用 LLM,并在浮标下方显示总分、覆盖度、维度分和评审来源;再次点击「打分」可收起/展开评分结果。
  • 评分结果下方提供「语义评审」按钮:执行 /score --semantic,先探测大模型 provider,成功后更新评分卡。
  • 浮标默认贴着右上角;直接拖动任意 Insight 或打分图标即可更换位置。

打开流程:

POST /api/dsh-insight/open { sessionId }
→ 发现运行中的 Insight
→ 增量导出/导入当前 DSH 会话
→ 浏览器打开 /session/<sid>?framework=dsh

如果 Insight 尚未准备好,浮窗和命令会等待同一个 Host 级任务,并显示 “检查 / 下载 / 启动”阶段。切换 Session 不会中断下载。启动 Insight 时设置了 CONTEXT_NO_BROWSER=1,因此不会先打开 Insight 首页;导入完成后只打开当前会话页。

如果下载失败,浮窗会显示 npm 错误,完整输出位于:

~/.dsh-insight/runtime/install.log
~/.dsh-insight/runtime/start.log

评分卡右侧边栏暂未注册;后续 UI 可以直接消费:

GET /api/dsh-score/scorecard?taskId=<sid>&cover=1