@cyberbouncer/mcp
v0.1.1
Published
MCP server for CyberBouncer — let Claude, Cursor, and other MCP clients run website security scans and read graded reports.
Maintainers
Readme
@cyberbouncer/mcp
MCP server for CyberBouncer — run website security scans from Claude, Cursor, or any MCP client. "We check IDs at the door."
Tools
- scan_url — passive security scan of any public URL: graded report (A–F, 0–100) covering TLS, security headers, cookies, email auth, DNS, and reputation, with prioritized findings.
- get_fix_prompt — scan plus a ready-to-apply remediation brief (prioritized fixes with config/code guidance) for your AI assistant to act on.
Setup
Claude Desktop / Claude Code
{
"mcpServers": {
"cyberbouncer": {
"command": "npx",
"args": ["-y", "@cyberbouncer/mcp"],
"env": {
"CYBERBOUNCER_API_KEY": "cbk_your_key_here"
}
}
}
}Cursor
Settings → MCP → Add server, command: npx -y @cyberbouncer/mcp.
API key (optional)
Works without a key (12 scans/min per IP). A free key from your
CyberBouncer dashboard raises that to
100 scans/day (paid plans: 1,000–5,000/day). Set it as
CYBERBOUNCER_API_KEY.
Env
| Variable | Default | Purpose |
| --- | --- | --- |
| CYBERBOUNCER_API_KEY | — | Your cbk_… API key |
| CYBERBOUNCER_API_URL | https://cyberbouncer.io/api | API base override |
Scans are passive only — no logins, no form submissions, no exploitation. See the API docs and methodology.
