Connect your agent to the Scan.now MCP server

Two anonymous, read-only tools for authorised passive website security scans and public phishing-link checks. No Scan.now account or API key is required.

Connection details

Streamable HTTP
https://mcp.scan.now

Add this URL as a remote Streamable HTTP MCP server in a compatible agent host. The endpoint is public and uses no authentication, so there is no API key to create, store or send.

Available tools

scan_website_security

Runs bounded, passive checks against a public website. The agent must first confirm that the user owns the target or has explicit permission to assess it.

  • url: the public HTTP or HTTPS website.
  • authorized: must be true only after the user confirms permission.
  • scope: optional; full, tls, headers, email or js.

check_phishing_link

Inspects a public link, follows its redirect chain when requested and reports phishing signals such as deceptive hosts or lookalike domains. It does not submit forms or log in.

  • url: the public link to inspect.
  • follow_redirects: optional; defaults to true.

Safety boundary

The tools do not exploit findings, bypass access controls, guess credentials or scan private networks. A clean scan is a point-in-time result, not proof that a site or link is safe.

Want the ready-made instructions?

The separate Scan.now CLI and agent skill page includes the complete SKILL.md and the one-command OpenClaw installation.

Privacy and methodology

See how Scan.now scans for the exact checks and limits. MCP-specific data handling is covered by the MCP privacy policy.