The domain tools speak the Model Context Protocol (MCP) and ship a REST API, so your AI assistant can check domain availability for you. Connect once, then just ask. Free and open source.
mcp_server.py) runs on your machine with nothing to host and works with every desktop and CLI client. Remote MCP (mcp_http.py or the Vercel function at /api/mcp) connects web assistants by URL.
One deploy hosts the website, the REST API, and the remote MCP connector on a single domain. The included vercel.json needs no changes.
https://webtoolshub-blond.vercel.app/api/mcp (remote MCP) and https://webtoolshub-blond.vercel.app/api/check (REST API).Add to claude_desktop_config.json, then restart.
{ "mcpServers": { "webtoolshub": { "command": "python", "args": ["C:/path/to/webtoolshub/mcp_server.py"] } } }
claude mcp add webtoolshub -- python /path/to/webtoolshub/mcp_server.py
Settings > Connectors > Add custom connector, then paste your deployed URL.
https://webtoolshub-blond.vercel.app/api/mcp
{ "tools": [ { "type": "mcp", "server_label": "webtoolshub", "server_url": "https://webtoolshub-blond.vercel.app/api/mcp" } ] }
In the ChatGPT app, add the same /api/mcp URL under Settings > Connectors. Or build a custom GPT Action by importing your hosted openapi.yaml.
{ "mcpServers": { "webtoolshub": { "command": "python", "args": ["/path/to/webtoolshub/mcp_server.py"] } } }
For a remote connection use "httpUrl": "https://webtoolshub-blond.vercel.app/api/mcp". The Gemini API can also call /api/check via function calling using openapi.yaml.
Any MCP client accepts the same local server config.
{ "mcpServers": { "webtoolshub": { "command": "python", "args": ["/path/to/webtoolshub/mcp_server.py"] } } }
One connector, one setup. Connect WebToolsHub once and your assistant gets every tool below through the same MCP server (/api/mcp) or REST API.
| Tool | Arguments | Returns |
|---|---|---|
check_domain | name, tlds? | Availability per TLD |
check_domains_bulk | names, tlds? | Availability per name and TLD |
whois_lookup | domain, include_raw? | Full WHOIS/RDAP record: registrar, dates, status, nameservers, DNSSEC |
dns_lookup | domain, types? | DNS records (A, AAAA, CNAME, MX, NS, TXT, SOA, CAA) |
ssl_check | host, port? | SSL/TLS certificate: issuer, expiry, SAN, TLS version, validity |
email_auth | domain, dkim_selector? | SPF, DMARC, DKIM and MX analysis with a grade |
who_hosts | domain | Hosting IP, ASN, org, country and reverse DNS |
dns_propagation | domain, type? | Same record across Google, Cloudflare, AdGuard, Quad9 |
find_subdomains | domain | Subdomains from Certificate Transparency logs |
security_headers | target | HTTP security-header grade with fixes |
trace_redirects | target | Full HTTP redirect chain, hop by hop |
portfolio_expiry | domains[] | Bulk domain expiry, sorted by urgency |
idn_homograph | domain | Punycode convert and lookalike / homograph detection |
REST: the first four have direct paths (/api/check, /api/whois, /api/dns, /api/tls); the rest run through one dispatcher, GET /api/tool?name=<tool>&..., e.g. /api/tool?name=email_auth&domain=github.com. Full schema in openapi.yaml. Every tool is served by the same MCP endpoint, so one connector gives your AI all twelve.
Which AI assistants work? Any that support MCP: Claude Desktop, Claude Code, Claude.ai, ChatGPT and OpenAI API, Gemini CLI and API, Kimi, Cursor and Windsurf. Anything that can call HTTP can use the REST API.
Do I need to host anything? Not for local clients. For web assistants that connect by URL, deploy to Vercel (free, hosts the whole project on one domain).
Is it free? Yes, free and open source under MIT.