How to Give Claude and Cursor Web Diagnostic Tools

Short answer. Connect Claude Desktop or Cursor to the enterno.io MCP server and the assistant gains 16 diagnostic tools: SSL, DNS, HTTP headers, Ping, WHOIS, a security scanner, and more. Setup takes a couple of minutes: add the URL https://enterno.io/mcp to your config, restart the client, and ask it to check a domain in plain text. No key is needed for public checks.

Why this beats copy-pasting

Without tools, an assistant can't run a real check — it either refuses or guesses. The usual workaround is running utilities by hand and pasting the output into chat. That's slow and goes stale fast. An MCP server gives the assistant direct access to tools, so it runs the checks itself whenever they're needed to answer.

With an MCP server connected, Claude stops hallucinating about headers and SSL — it actually checks them and answers from facts.

Step 1. Open your client config

• Claude Desktop — the claude_desktop_config.json file (there's an "Edit Config" button in app settings).
• Cursor — the mcp.json file in your project directory or in global settings.
• Zed — the MCP settings section in the editor config.

Step 2. Add the server

Insert the block with the server URL. It's identical across all clients that support remote MCP over streamable HTTP:

{
  "mcpServers": {
    "enterno": {
      "url": "https://enterno.io/mcp"
    }
  }
}

Save the file and restart the client. After the restart, the server's tools appear in the list available to the assistant.

Step 3. Verify the tools connected

Ask the assistant which tools it has, or just hand it a task. If the server is connected, the response will include tool calls rather than invented text.

Check the SSL chain and DNS for example.com
and tell me if the security headers are fine.

This prompt makes the assistant call the SSL, DNS, and security tools and merge the results into a single analysis.

What the assistant actually gets

That's 16 tools in total. For public checks against third-party domains, no key is required — the server is free.

Programmatic access via the SDK

If you're building your own agent instead of using a ready client, install the SDK:

npm install @enterno/mcp-client

The @enterno/mcp-client package gives typed access to the same 16 tools over streamable HTTP — handy for scripts and CI integration.

Security rule: only give your agent servers you trust, and review the tool list before connecting. The enterno.io diagnostic tools make outbound checks only and never touch your data.

Common setup mistakes

• Forgot to restart the client — config changes apply only after a restart.
• Broken JSON — a stray comma or quote breaks the whole file; validate it.
• Wrong config path — every client has its own file; make sure you're editing the right one.

What to try after connecting

Ask the assistant to check your site's HTTP headers, assess API security, or run an AI-readiness check. The assistant runs the real checks and explains the findings.

FAQ

Does this work only in Claude?

No. MCP is an open standard: Claude Desktop, Cursor, Zed, and other compatible clients connect the same way.

Do I need to install anything?

For a ready client — no, just the URL in the config. The @enterno/mcp-client SDK installs via npm if you're writing your own agent.

How much does it cost?

Public checks are free and keyless. On top of that, enterno.io offers 10 free monitors and 48+ free on-site tools.

Is it safe to connect?

The diagnostic tools make outbound checks against third-party URLs only, with SSRF protection. They have no access to your code or data.

Can I use it in CI?

Yes — via the @enterno/mcp-client SDK, call the tools from scripts, for example to check SSL before a deploy.

Connect the enterno.io MCP server →