Web Utilities

Web Utilities

Generic primitives for asking simple questions about a public URL. No outside account required.

Capabilities

• DNS resolve: Given a hostname, return whether it resolves to an IPv4 address. Use as a cheap first-pass liveness check before a full HTTP fetch — fabricated or expired domains return alive=false.
• HEAD probe: Given a URL, send an HTTP HEAD request that follows redirects. Returns final URL, status code, content type, and an alive flag (true for 2xx/3xx/4xx, false for 5xx, network failure, or timeout). Faster than a full GET when only "is this reachable" matters.
• Page metadata fetch: Given a URL, fetch the response body and parse the HTML for <title>, <meta name=description>, og:description, the first <h1>, and the first <p>. Returns whichever fields the page actually exposes. Truncates each field at a reasonable length so output stays compact for downstream agents.

Operational notes

• All three tools run with strict timeouts (default 5 seconds) and follow redirects.
• HEAD probe falls back to GET when the server returns 405 Method Not Allowed for HEAD; the response body is discarded so the cost stays low.
• Metadata fetch only parses text/html responses. PDFs, JSON endpoints, and other non-HTML content return empty metadata fields with a status code so the caller can distinguish "no metadata" from "page broken".
• Errors that prevent the request from completing (DNS failure, connection refused, TLS failure, timeout) are returned as alive=false with a short reason string. Upstream HTTP error codes (4xx/5xx) are returned with the status code intact.