# Sinewaveai Agent Security Scanner MCP server

Source: https://glimind.com/server/mcp-registry%2Fio.github.sinewaveai%2Fagent-security-scanner-mcp
Registry id: `mcp-registry/io.github.sinewaveai/agent-security-scanner-mcp`
Status: unknown

## Install

```json
{
  "mcpServers": {
    "sinewaveai-agent-security-scanner-mcp-se": {
      "command": "npx",
      "args": [
        "-y",
        "agent-security-scanner-mcp"
      ]
    }
  }
}
```

Claude Code:

```sh
claude mcp add sinewaveai-agent-security-scanner-mcp-se -- npx -y agent-security-scanner-mcp
```

## Tools (10)

- `check-package`: Check a single package before installing it.
- `demo`: Run a demo scan with tool poisoning and command-exec findings.
- `doctor`: Run a local environment health check.
- `init`: Install the scanner into an AI coding client (e.g., claude-code, cursor).
- `init-ci`: Add a GitHub Actions workflow for CI scanning.
- `quickstart`: Get repo-specific next steps for a given AI client.
- `scan-mcp`: Audit an MCP server for vulnerabilities before adding it to an AI client.
- `scan-packages`: Verify AI-suggested imports are real packages, not hallucinations.
- `scan-project`: Scan a project for security issues and get an A-F security grade.
- `share-kit`: Generate public-safe copy for launch posts, GitHub issues, or directory listings.

Measured by Glimind (https://glimind.com): MCP handshake + tool list only, no tool is ever called. CC BY 4.0.