# 使用 Homebrew 安装 snyk-agent-scan

查看 snyk-agent-scan 的安装路径、可执行文件、元数据以及面向 AI 代理工作流的安全说明。

## 安装

```sh
sudo av install brew:snyk-agent-scan
```

其他安装命令:

### macOS

- Homebrew (100%):

```sh
brew install snyk-agent-scan
```

  证据: local Homebrew formula metadata

## 软件包事实

- **软件包键:** brew:snyk-agent-scan
- **软件包管理器:** Homebrew
- **版本:** 0.5.15
- **来源摘要:** Constrain, log and scan your MCP connections for security vulnerabilities
- **主页:** <https://github.com/snyk/agent-scan>
- **最后更新:** 2026-07-17T17:13:21Z
- **已生成:** 2026-08-03T19:37:03+00:00

## 可执行文件

- snyk-agent-scan (别名)

## 安装行为

- Bottle: 不可用

## 版本和新鲜度

- 页面生成时间: 2026-08-03
- 管理器版本: 0.5.15
## 项目历史与用法

Snyk Agent Scan is Snyk's CLI for discovering and scanning local AI agent components, MCP servers, skills, and related configuration for security risks. It reflects the new package-manager problem of executable agent ecosystems: installed tools are no longer just binaries, they are prompts, skills, MCP configs, and commands that may run during inspection.

### 项目历史

The public GitHub repository was created in 2025. Its README describes Agent Scan as a tool for inventorying installed agent components and scanning for prompt injections, sensitive data handling, malware payloads hidden in natural language, tool poisoning, toxic flows, and vulnerabilities in agent skills.

The README notes that CLI output is experimental and that Agent Scan 0.4 was published with a technical report on emerging threats in the agent skill ecosystem, indicating a young tool aimed at a fast-changing area.

### 采用历史

Official usage examples install and run the tool with `uvx snyk-agent-scan@latest`, while the input records a Homebrew formula. The supported-agent matrix includes mainstream developer agents such as Claude, Cursor, Windsurf, Gemini CLI, VS Code, Codex, Amazon Q, and others.

### 使用方式

Users can run a full machine scan, scan a specific MCP configuration file, scan one skill file, or scan a directory of skills. The README warns that scanning MCP configurations may execute commands defined in those configs, so the CLI includes interactive consent and a deliberately named flag for trusted non-interactive runs.

### 为什么软件包爱好者会关心

Package nerds care because Agent Scan treats local agent configuration as supply chain surface. It scans the places modern developer tools install behavior, including project, user, system, extension, and plugin scopes, making it relevant to anyone packaging or auditing AI-enabled command-line environments.

### 时间线

- 2025: Public GitHub repository created.
- 2025: README documents scanning for MCP servers, agent tools, and skills.
- 2026: Repository metadata shows active maintenance.

### Related projects

- Related ecosystems include MCP servers, Codex skills, Claude skills, Cursor/Windsurf/Gemini agent configurations, and Snyk's broader developer security tooling.

### 来源

- <https://github.com/snyk/agent-scan>
- <https://github.com/snyk/agent-scan#readme>
- source_facts.package-manager


## 安全说明

没有找到 snyk-agent-scan 的匹配本地密钥处理 manifest。Nucleus 软件包元数据仍在此发布，以便未来覆盖拥有稳定的软件包 URL。



## Combined YAML source

View the package source record on GitHub. [combined/snyk-agent-scan.yml](https://github.com/mxcl/pkgdb/blob/main/combined/snyk-agent-scan.yml)


## 来源

- pkg.so package database
- Geiger risk classifier
- curated package history
- pkgdb category and tag curation
- cross-ecosystem install command graph