Acceptable use
Acceptable Use for Webpage Extraction
WEB/TEXT converts one public HTTPS page within URL, DNS, redirect, script-blocking, size, timeout, and rate-limit boundaries. It is not a bypass tool, credentialed browser, private-network client, or unrestricted content mirror.
Use publicly accessible pages only
WEB/TEXT accepts one public HTTPS page that you can reach without signing in or supplying private access material. It does not promise access to pages limited by login walls, paywalls, CAPTCHAs, geographic controls, robots directives, or other technical barriers.
A failed or blocked request remains a failed or blocked request. The product explains the boundary instead of cycling through unbounded retries or presenting a page shell as a complete document.
Never submit credentials or private files
Do not provide cookies, passwords, session tokens, authorization headers, private document links, or account-only URLs. WEB/TEXT does not request those values or attempt to inherit them from your browser.
Private dashboards, inboxes, cloud drives, internal tools, and local files require different authorization and security models and remain outside this workflow.
Validate every URL before a public fetch
URL validation only accepts public HTTPS URLs on the standard HTTPS port. It rejects embedded credentials, localhost, private-network ranges, link-local addresses, cloud-metadata endpoints, and unsupported schemes before the Worker fetches HTML.
The service resolves the submitted hostname through DNS, follows redirects manually, and repeats URL and DNS validation for each redirect before fetching the next hop. Redirect loops, excess hops, and non-public answers fail closed.
Block scripts and subresources
Browser Run receives bounded temporary HTML instead of the target URL. JavaScript is disabled, and scripts and subresources are rejected rather than loaded from the source page. Pages whose readable content exists only after JavaScript runs may fail.
Responses use bounded request and result sizes, browser timeouts, and client and target rate limits. A page outside those boundaries returns a failure instead of consuming resources indefinitely.
Keep the single-URL workflow within scope
The public converter handles one user-requested URL at a time. Do not use it to disguise an unbounded crawl, enumerate an entire site, or distribute large-scale extraction across repeated browser sessions.
Batch processing and site crawling need separate limits, ownership checks, queue controls, observability, and commercial terms. They are not part of this public converter.
Keep output attributable and temporary
Extracted output keeps source context so you can verify where it came from. Removing page chrome does not make WEB/TEXT the owner of the underlying content.
WEB/TEXT does not persist extracted page bodies. Temporary copy and download actions remain under your control; a public result library, permanent archive, or shared index is not part of the service.