GitHits
Search public open-source code, documentation, metadata, vulnerabilities, changelogs, and examples.
Security check for coding agents: malware, supply chain, secrets and web bugs. Reads, never runs.
列表按使用分(满分 100)排序,由以下部分组成
| 热度 | 6/80 | GitHub 星标或每周下载量:较大的计 60,较小的计 20 |
|---|---|---|
| 新鲜度 | 12/12 | 30 天内有发布或代码更新得满分 |
| 仍在响应 | 0/8 | 上次检查时云端端点有响应 |
| 使用分 | 18/100 |
质量: 基础检查通过 3/3 项(仅展示,不用于排序) · 上升势头: 暂无数据 (只用于“上升中”标签)
选择你使用的 AI 应用,然后复制设置。
通过 3/3 项检查 · 在你的电脑上运行
第一步:你用哪个 AI?
选择你用的应用,步骤会显示在下方。
可连接 Claude,但需要先准备
不需要密钥,不用安装程序 · 仅限电脑上的 Claude 应用(Windows 或 Mac),网页和手机不可用
用 vetcode 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Claude 说正在使用 vetcode,或回答中出现 vetcode 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Anthropic 帮助页面核实 · 你的界面可能因版本略有不同
ChatGPT 暂时无法连接 vetcode
vetcode 需要安装在你的电脑上,而 ChatGPT 只能连接在网上运行的服务器
A · 使用能连接 vetcode 的应用
B · 网络安全 类中能连接 ChatGPT 的其他服务器
从同类中使用较多的里挑选,仅限用地址即可连接、无需密钥的
Claude Code 暂时无法连接 vetcode
开发者只提供了电脑版 Claude 应用的安装文件
A · 使用能连接 vetcode 的应用
Cursor 暂时无法连接 vetcode
开发者只提供了电脑版 Claude 应用的安装文件
A · 使用能连接 vetcode 的应用
VS Code 暂时无法连接 vetcode
开发者只提供了电脑版 Claude 应用的安装文件
A · 使用能连接 vetcode 的应用
依据这个服务器的连接方式和各应用自己的文档,检查于 2026年11月10日。
在你的电脑上运行,适用于
ClaudeGooseLM StudioMsty StudioCherry StudioVisual StudioWindsurf (Devin Desktop)ZedKiroClineContinueJetBrains AI AssistantCodex CLIGemini CLIOpenCodenodecodeWarpLibreChatRaycast
它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月16日
注册名称已通过 GitHub 账户 openyard-cloud 验证,该账户拥有此仓库
代码最后更新于3天前
使用 MIT
它在你的电脑上运行,没有互联网端点
在你的电脑上运行,改用本地设置
它在你的电脑上运行;我们只能读取互联网端点的工具列表
如果你是作者,可把这段代码贴到 README。徽章会链接回本页。
[](https://mcp.popnx.com/en/servers/openyard-cloud-vetcode)提问、分享技巧或报告无法使用。按“有帮助”数量排序。
还没有评价
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 3 项适用的基础检查中通过了 3 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。
GitHub 仓库最后更新于3天前,最新版本(0.8.7)于3天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 MCPB。
同一分类中有:GitHits, Dependency Management MCP Server, Snyk MCP, Draugr, Spotdb
同一分类、其他作者、命令已通过检查。官方的排在前面,再按星标排序。
如何看卡片:↓ 每周下载量i★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
Search public open-source code, documentation, metadata, vulnerabilities, changelogs, and examples.
Sonatype component intelligence: versions, security analysis, and Trust Score recommendations
Easily find and fix security issues in your applications leveraging Snyk platform capabilities.
Security scanning for AI agents: SAST, SCA, secrets, IaC, DAST, ranked by real risk.
Ephemeral data sandbox for AI workflows with guardrails and security
| 本项 vetcode | GitHits | Dependency Management MCP Server | Snyk MCP | Draugr | Spotdb | |
|---|---|---|---|---|---|---|
| 运行位置 | 需要安装在电脑上 | 云端或安装在电脑上都可以 | 在云端运行,无需安装 | 需要安装在电脑上 | 需要安装在电脑上 | 需要安装在电脑上 |
| 需要密钥 | 否 | 否 | 否 | 否 | 否 | 否 |
| 可在 Claude 中按 + 添加 | 否 | 否 | 否 | 否 | 否 | 否 |
| 基础检查 | 3/3 | 5/5 | 3/4 | 3/3 | 3/3 | 3/3 |
| 许可证 | MIT | Apache-2.0 | – | Apache-2.0 | Apache-2.0 | MIT |
| 价格 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 |
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
vetcode
The security check your AI coding agent runs before it acts. Before it installs a package, after it clones a repository, before your code ships. Zero dependencies. Reads, never runs. Works without an agent too.
# Claude Code: the skill plus the guard that checks installs and clones
claude plugin marketplace add openyard-cloud/vetcode && claude plugin install vetcode@openyard
npx skills add openyard-cloud/vetcode # any other agent, as a skill npx github:openyard-cloud/vetcode#v0.8.7 . # no agent: scan a project, a package or a cloned repo
An agent installs a package in a second and runs a cloned project without reading it. vetcode is the check in between. It is three things:
| What it does | Where it runs | |
|---|---|---|
| The guard | Answers allow, ask or deny before a package is installed or an MCP server is added, and reads a repository right after it is cloned | Claude Code, on its own; other agents call it as a tool |
| The scanner | Reads a folder for malware, supply-chain attacks, attacks on agents, leaked secrets, and server and CI problems, and lists the places in web code worth reading | Skill, MCP server, command line, GitHub Action |
| The review procedure | Tells your agent how to confirm each finding by reading the code, and what to report | Skill, MCP server |
Nothing from the code it looks at is installed, built or run. It needs Node.js 18 or newer and nothing else. Free and open source (MIT), from the OpenYard community.
Contents: How good is it · Install · The guard · Use it · What it checks · How it works, and its limits
Every number here is measured, on samples the rules were not tuned on unless it says so. The scripts, the runs and the full history are in bench/.
| Question | vetcode | For comparison |
|---|---|---|
| Malicious npm packages caught, of 1,000 (false alarms on 1,000 clean ones) | 947 (26); verdicts 907 (5) | GuardDog: 903 (43) |
| Malicious PyPI packages caught, of 300 (false alarms on 210 clean ones) | 262 (16); verdicts 234 (1) | GuardDog: 241 (48) |
| The same malware after 8 kinds of disguise | still catches 99–100% of what it caught | GuardDog: 90–92% on several |
| Malicious agent skills it had never seen, of 157 (false alarms on 1,029 clean ones) | 91 (1) | Cisco skill-scanner: 101 (56) |
| The guard, shown malware as if nobody had reported it yet | blocks or questions 95% of npm, 88% of PyPI, and 3% of clean npm packages | |
| A package whose newest release is hours old, asked for 3 times | the agent installs the previous, settled version 3 times | the agent alone installs the new release 3 times |
| Wrong "this is malicious" verdicts on 72 real applications (228,000 files) | 1 | |
| Findings an agent has to read on those applications | about 4 per 1,000 files, plus 8 places to read | |
| Real injection bugs the scanner points at, of 480 advisories | 55 (11%); 6% on the half no rule was tuned on | |
| Real authorization bugs found by Claude Code with vetcode, of 49 advisories | 41% | Claude Code alone does as well |
A sign and a verdict are different things. A scanner of several hundred rules that lets any one of them call code malicious is wrong somewhere on almost every large project. vetcode reports every sign, which is what your agent is told to read, and gives a verdict, which is what stops an install without asking, only when a rule with a measured record fires, when two rules agree on one file, or when the sign sits in code an install script starts.
What it is good at. Malware and supply-chain attacks in packages and repositories, attacks aimed at coding agents (poisoned CLAUDE.md and skills, hooks and auto-approved MCP servers in project settings, tool descriptions that give orders), leaked secrets, and stepping in before the install, not after.
What it is not. It does not find web bugs by itself: it lists the dangerous calls as places to read, and those stand at about one real injection bug in ten, so they are where to start, not coverage. It does not make your agent better at logic bugs such as missing authorization; that is the agent's own ability. And a clean report means "nothing obvious found", not "safe".
Read the numbers with care. Samples are 157 to 1,000 items, so they move by a point or two between samples. Most malicious packages in the public datasets were found by GuardDog itself, which favours it. Cisco's scanner catches ten more skills by flagging 56 clean ones, a trade vetcode does not make. All numbers are the scanner alone; in normal use your agent reads every finding again.
All forms need Node.js 18+. vetcode is not on npm: every command here runs it from this repository, pinned to a release. A package called vetcode on npm or PyPI is not this project.
| Your agent | Use |
|---|---|
| Claude Code | the plugin (skill and guard), or the skill alone |
| Gemini CLI | the extension (skill, guard and MCP server) |
| Codex, GitHub Copilot, Cursor, Amp, Goose, opencode, Windsurf, Cline, Roo… | the skill |
| Anything that speaks MCP (Cursor, VS Code, Codex, Gemini CLI, Windsurf, Zed, Claude Desktop, ChatGPT…) | the MCP server |
| claude.ai | the skill zip |
| CI, scripts, or no agent at all | the command line or the GitHub Action |
claude plugin marketplace add openyard-cloud/vetcode && claude plugin install vetcode@openyard
Inside a session the same is two commands, because Claude Code only installs from marketplaces it already knows:
/plugin marketplace add openyard-cloud/vetcode
/plugin install vetcode@openyard
For a whole team
Commit this to the project's .claude/settings.json. Everyone who opens the project and trusts the folder gets the marketplace registered, so each person only runs /plugin install vetcode@openyard once:
{
"extraKnownMarketplaces": { "openyard": { "source": { "source": "github", "repo": "openyard-cloud/vetcode" } } }, "enabledPlugins": { "vetcode@openyard": true } }
The skill, the guard and the MCP server in one install:
gemini extensions install https://github.com/openyard-cloud/vetcode
Gemini CLI downloads the extension from the latest release. Its guard runs before every shell command and file write, like the Claude Code one, with one difference: Gemini CLI has no "ask" answer for a hook, so where Claude Code would ask, you see vetcode's warning and then Gemini CLI's own confirmation of the command. With auto-approval on (--yolo), set VETCODE_GUARD=strict to turn those warnings into a refusal.
For any agent that loads Agent Skills. The skills CLI knows where each agent keeps them (vetcode on skills.sh):
npx skills add openyard-cloud/vetcode
Or copy the folder yourself
git clone https://github.com/openyard-cloud/vetcode, then copy vetcode/skills/vetcode into your agent's skills folder, in the project or under your home folder: .agents/skills/ (shared by many agents, such as Codex and Copilot), .claude/skills/ (Claude Code), .github/skills/ (Copilot), .gemini/skills/ (Gemini CLI) or .cursor/skills/ (Cursor).
Four tools (vetcode_scan, vetcode_guard, vetcode_guide, vetcode_reference) and a security_review prompt. Most agents (.cursor/mcp.json, .mcp.json, Gemini CLI settings.json, Windsurf mcp_config.json):
{
"mcpServers": { "vetcode": { "command": "npx", "args": ["-y", "-p", "github:openyard-cloud/vetcode#v0.8.7", "vetcode-mcp"] } } }
Claude Desktop: download vetcode.mcpb from Releases and open it. The server is also in the MCP Registry as io.github.openyard-cloud/vetcode.
VS Code and Codex
VS Code (.vscode/mcp.json):
{ "servers": { "vetcode": { "type": "stdio", "command": "npx", "args": ["-y", "-p", "github:openyard-cloud/vetcode#v0.8.7", "vetcode-mcp"] } } }
Codex (~/.codex/config.toml):
[mcp_servers.vetcode]
command = "npx" args = ["-y", "-p", "github:openyard-cloud/vetcode#v0.8.7", "vetcode-mcp"]
Then ask your agent to "use vetcode to review this project". The agent should call vetcode_guide once; the server also tells it to.
Download vetcode-skill.zip from Releases and upload it under Settings → Capabilities → Skills.
For tools that read AGENTS.md (or .cursorrules, CONVENTIONS.md…) but neither skills nor MCP, clone this repository next to your project and add:
## Security reviews
When asked to check code for security problems or malware, follow ../vetcode/skills/vetcode/SKILL.md and run its scanner: node ../vetcode/skills/vetcode/scripts/vet.mjs .
An agent decides to install a package in a second: a name it remembered wrong, a name a web page suggested, a release published an hour ago. Installing runs that package's code with your rights. With the Claude Code plugin or the Gemini CLI extension, vetcode looks at every install command before it runs, and at every MCP server before it is registered, and answers allow, ask or deny:
| The command would… | Answer |
|---|---|
| install a package that does not exist (a made-up name is one an attacker can register) | deny |
| install a package with a malware advisory on OSV.dev | deny |
| install a package whose code gets a malware verdict at critical (the archive is downloaded and read, never run) | deny |
| install a package whose code shows signs of malware short of that | ask |
install a name one letter from a popular package (expres, reqeusts) that few people install | ask |
| install a release less than 72 hours old, or one that lost its provenance or gained an install script | ask |
| pipe a download into a shell, or install from another index, a repository or a URL | ask |
| add an MCP server at a tunnel address (ngrok, trycloudflare…) or over plain HTTP to another machine | ask |
| install well-known packages at settled versions | allow, silently |
"Ask" shows you the reason and waits fo