toolhq.io

mcp

Give your AI agent real network diagnostics.

Connect toolhq to a compatible AI coding tool and let your agent inspect DNS, TLS, email and network infrastructure without leaving your workflow. No key, no account, one endpoint.

toolhq — mcp
$ claude mcp add --transport http toolhq https://toolhq.io/api/mcp

what your agent can do

These are the checks that need real network access, the thing a model cannot do on its own. Every one runs the same code as the tool on the site.

dns_lookup

Read every DNS record for a domain, or one record type at a time.

DNS Lookup in the browser
ssl_check

Inspect a host’s TLS certificate: validity, issuer, SAN, chain, and expiry.

SSL Checker in the browser
mx_lookup

List a domain’s mail servers with priorities and resolved addresses.

MX Lookup in the browser
ip_lookup

Describe and geolocate a public IPv4 or IPv6 address.

IP Address Lookup in the browser
ip_reputation

Tell whether an address is a VPN, proxy, Tor exit, or datacenter host.

VPN & Proxy Detector in the browser
http_headers

Fetch a URL and report its status, redirect chain, and response headers.

HTTP Header Checker in the browser
dns_propagation

Compare a record across six public resolvers to see if a change has landed.

DNS Propagation Checker in the browser
reverse_dns

Resolve PTR records for an address, with forward confirmation.

Reverse DNS (PTR Lookup) in the browser

things to ask for

connect your client

Pick your agent and paste the configuration. The server speaks the MCP Streamable HTTP transport, so any client that supports a remote HTTP MCP server will work even if it is not listed here.

Claude CodeTerminal
claude mcp add --transport http toolhq https://toolhq.io/api/mcp

privacy and limits

questions

Does toolhq MCP need an API key or an account?

No. The endpoint is open over HTTP and needs no key, token or account. Requests are rate limited by IP address.

Which checks are available to an agent?

8: dns_lookup, ssl_check, mx_lookup, ip_lookup, ip_reputation, http_headers, dns_propagation, reverse_dns. These are the checks that need real network access, which is exactly what a model cannot do on its own. The browser-only utilities such as the JSON formatter and JWT decoder are not exposed, because an agent can already do that work locally.

What does toolhq see when my agent runs a check?

The argument the agent sends: a domain, a host, an IP address or a URL. That value is used to make the lookup and is not stored. The same SSRF guards and rate limits as the website apply.

Is it stable?

The endpoint is in beta. The transport is the MCP Streamable HTTP transport and the tool names are stable, but new checks may be added and responses may gain fields.

Browse the tools