← WebMCP Directory

Tool tree

scanwebmcp.com

Developer ToolsTools verified
Website
A published, versioned rubric for how much of a business an AI agent can see, understand, and act on.
Explore capabilities

    page: Every page

    scan_agent_surfaceAction

    Scan a website with ScanWebMCP.com. Returns its public rung and scores. In a scan-and-email request, call this first and email_report only after it succeeds. Takes 10–30 seconds.

    View tool JSON
    {
      "name": "scan_agent_surface",
      "kind": "act",
      "impl": "imperative",
      "description": "Scan a website with ScanWebMCP.com. Returns its public rung and scores. In a scan-and-email request, call this first and email_report only after it succeeds. Takes 10–30 seconds.",
      "inputSchema": {
        "type": "object",
        "properties": {
          "url": {
            "type": "string",
            "description": "The website to scan, e.g. example.com"
          }
        },
        "required": [
          "url"
        ]
      },
      "page": "/ladder"
    }
    get_ladder_definitionAnswer

    The Agent Surface Ladder v1.0 — the published rubric this site scores against. Optionally pass a rung number (0–4) for one rung's definition.

    View tool JSON
    {
      "name": "get_ladder_definition",
      "kind": "answer",
      "impl": "imperative",
      "description": "The Agent Surface Ladder v1.0 — the published rubric this site scores against. Optionally pass a rung number (0–4) for one rung's definition.",
      "inputSchema": {
        "type": "object",
        "properties": {
          "rung": {
            "type": "number",
            "description": "Optional rung 0-4"
          }
        }
      },
      "page": "/ladder"
    }
    email_reportSensitive Action

    Send the full evidenced ScanWebMCP.com report for an already-scanned website to an email address (run scan_agent_surface first if the site hasn't been scanned). CONSEQUENTIAL — sends one transactional report. Benchmark updates require a separate boolean opt-in and email confirmation. A direct request to send this report to an address the human supplied is sufficient confirmation for the transactional email. Never guess, look up, or auto-fill an address.

    View tool JSON
    {
      "name": "email_report",
      "kind": "transact",
      "impl": "imperative",
      "description": "Send the full evidenced ScanWebMCP.com report for an already-scanned website to an email address (run scan_agent_surface first if the site hasn't been scanned). CONSEQUENTIAL — sends one transactional report. Benchmark updates require a separate boolean opt-in and email confirmation. A direct request to send this report to an address the human supplied is sufficient confirmation for the transactional email. Never guess, look up, or auto-fill an address.",
      "inputSchema": {
        "type": "object",
        "properties": {
          "email": {
            "type": "string",
            "description": "The address the human directly requested for this report."
          },
          "url": {
            "type": "string",
            "description": "The scanned website the report is for, e.g. example.com."
          },
          "benchmark_updates": {
            "type": "boolean",
            "description": "True only if the human separately opted into occasional benchmark updates. Defaults to false."
          }
        },
        "required": [
          "email",
          "url"
        ]
      },
      "page": "/ladder"
    }