Privacy approach
Privacy Approach for URL Extraction
WEB/TEXT keeps extraction results temporary: the target site sees a Cloudflare fetch, while Browser Run receives only bounded temporary HTML and not the submitted target URL as a navigation input.
What WEB/TEXT processes
When you submit a public HTTPS page, a Cloudflare Worker validates the URL and each redirect, then fetches the page through the public internet. The target site can see that Cloudflare fetch and the requested public URL. The service does not ask for login credentials, browser cookies, Authorization, custom headers, secrets, or private files.
WEB/TEXT does not persist extracted page bodies, create a public result page, or offer saved conversion history. Copy and download actions are available only from the temporary result in your current browser session.
Cloudflare Browser Run data flow
Browser Run receives only bounded temporary HTML after the Worker fetch. It does not receive the submitted target URL as a navigation input, browser cookies, Authorization, custom headers, or secrets. The supplied HTML can still contain text and links published by the target site.
Source scripts and subresources are blocked. Quick Actions handle the Markdown output temporarily, and default generated results can be cached for up to five seconds.
Keep submitted URL details in context
A publicly reachable page can still contain query parameters, campaign identifiers, document tokens, or personal information in its path. Avoid submitting a URL when you do not have a reason to send those details through WEB/TEXT's Cloudflare Worker to the target site.
The current result surface shows useful source context, such as domain, path, word count, and reading time. It does not turn that result into a public archive or a reusable content library.
Analytics and operational records
WEB/TEXT's routine analytics are configured to use product events and page paths rather than submitted query strings. Operational handling may still include limited request metadata needed to apply security controls, diagnose failures, and enforce rate limits.
Provider retention, logging, training, region, subprocessors, and deletion behavior are separate from WEB/TEXT's own storage choices. The Cloudflare fetch and temporary HTML-conversion flow described above are part of the current service boundary.
Use the public-page boundary
The extractor is for pages you can reach without signing in. Do not submit account-only URLs, private dashboards, cloud drives, inboxes, local files, or URLs that require cookies or credentials.
If a source changes its access controls or blocks automation, the converter may fail. It does not use credentials or anti-bot workarounds to obtain a result.