
3D code dependency graph generator with built-in LLM agent. Language-agnostic (Python, TypeScript, Rust, Go, Java, C/C++, C#, Ruby, Kotlin, Swift, PHP, Lua). Coupling depth analysis, constraint gating, real-time file watching. Tauri 2 + Three.js + Rust engine.跨语言代码依赖拓扑图生成器 · 14 门语言统一 IR · 3D 全息星图 · 内置 AI Agent 双向联动 · 四级耦合诊断 · 桌面应用 / CLI 双模
Install HoloGram
Clean install: prebuilt on npm and declares a `dsh.bundle` layer. Heads up: it runs install-time scripts (install) on your machine.
The one-line version
dsh plugin --profile my-profile add @a834063245/hologram-dshThe easy path: this bundle is published to npm, so the install pulls a prebuilt artifact — no build step runs on your machine and no build approval is needed.
From npm — installs a prebuilt artifact, no build step on your machine
# 1. add the bundle to a profile — --profile my-profile points at $DSH_HOME/profiles/my-profile
dsh plugin --profile my-profile add @a834063245/hologram-dsh
# 2. verify: the bundle should show up as a layer in the resolved config
dsh --profile my-profile --dump-config
# 3. boot the profile
dsh --profile my-profilePackage: @a834063245/hologram-dsh
Prefer to install from source instead?
From GitHub — pulls source, and may ask for build approval on first run
# 1. add the bundle to a profile — --profile my-profile points at $DSH_HOME/profiles/my-profile
dsh plugin --profile my-profile add github:834063245-creator/HoloGram#<commit-sha>
# 2. verify: the bundle should show up as a layer in the resolved config
dsh --profile my-profile --dump-config
# 3. boot the profile
dsh --profile my-profileBefore you run itinstall scripts
- This package declares
install— package-manager lifecycle scripts that execute on your machine at install time, outside the agent sandbox, before any tool-approval prompt exists. - A plugin runs with your permissions once loaded: it can read your files, use your credentials and reach the network. Tool approvals do not sandbox it.
Prerequisites
- API key
- Not required.
- Network access
- Not required at runtime.
- Language
- TypeScript
Install check
Every field the verdict was derived from, so you can re-derive it yourself
| Field checked | Result | Weight | What the spec says about it |
|---|---|---|---|
| lifecycleScripts | fail | warn | runs at install time, outside the agent sandbox: install |
| dsh.bundle | pass | fatal | package.json declares a `dsh.bundle` layer |
| dsh.bundle.patch file | pass | fatal | patch file present at `dsh-bundle/cordis.patch.yml` |
| npm registry | pass | fatal | published as `@a834063245/hologram-dsh` — prebuilt artifacts |
- Verdict reason
- on npm with a valid bundle declaration
- package.json read
- dsh-bundle/package.json
- Rules applied from
- docs/user/develop/basic/publish.md
- Checked at
- 2026-08-15
- Install-time scripts
install
From the README
Excerpt as published by the author, plain text, unedited
HoloGram 把代码库编译成一张统一 IR 依赖图(节点=符号/函数/类/模块,边=调用/继承/读写/时序),并通过 MCP 协议向 AI Agent 暴露 34 个图查询工具。 **核心主张:依赖推理应当是确定性的,而不是猜的。** LLM 分析"改 A 会炸什么"时,靠逐文件读源码推测依赖——弱模型会漏,大项目会翻不动。HoloGram 用 tree-sitter 静态分析预先算好整张依赖图:Agent 一次工具调用拿到结构化事实(影响面、循环、脆弱模块、数据流路径),而不是源文件文本。单点查询省 ~70% token,全局分析省 90%+;省 token 是次要的,**可靠性**是主要的。 引擎是单文件二进制,本地运行、零配置、代码不出机器。 引擎随 [Releases](https://github.com/834063245-creator/HoloGram/releases) 发布(Windows / Linux),安装脚本一键完成。也可以把下面这段话直接发给你的 AI 编程工具,让它自己装:
Topics
GitHub topics on this repository
More in Dsh Plugin Verify Browser Automation
Ranked by similarity inside the cluster, not alphabetically
Dsh Plugin
Tabbit-Browser
Tabbit Broser plugins for Deepseek Harness
Deepseek Harness Model Config
MarvekG
No description provided.
Dsh Playwright Browser
Clizo1209
Playwright browser automation for DeepSeek Harness|面向 DeepSeek Harness 的 Playwright 浏览器自动化插件
Dsh Image Bridge
kbpoyo
DSH 插件:让纯文本模型也能看图。Web 端直接粘贴图片即可发送,无需指定图片路径;模型自主调用视觉技能查看,多模态模型原生直通,零skill绑定。
Dsh Browser Playwright
ChenyuHeee
Snapshot-first Playwright browser automation for DeepSeek Harness: accessibility-tree interaction with stable refs, per-session browser contexts, 17 browser_* tools.
DeepSeek Harness Skill
itmoqing
这是一个Codex/Claude来进行任务发布给DeepSeek Harness干活的工作流的Skill,能实现并发,多个工作区一起执行