Link preview bot

Bot identity

Bot name Zirooms Link Preview

User-Agent Zirooms-LinkPreview/1.0

Public outbound IP list

Purpose

The bot collects titles, descriptions, preview images, and supported video information from public URLs that Zirooms and shotsuda members share in posts or comments. It runs in response to requests to write or view content. It does not crawl entire websites or run scheduled crawls.

Requests and limits

HTML, images, and official oEmbed information are fetched with GET requests. Video files are checked with HEAD requests or Range GET requests that limit how much data is downloaded.

The bot only requests public HTTP or HTTPS addresses. It does not forward users' cookies or credentials, or execute JavaScript on the page. It does not bypass login or security verification. If information cannot be fetched, the original link remains an ordinary link.

Each external lookup is limited to 5 seconds and up to 2 redirects. The bot reads up to 512KB of HTML and 1MB per image.

Caching and storage

Successfully collected general metadata is cached in memory for up to 24 hours, video information for up to 5 minutes, and images for up to 1 hour. Failed results are not cached. A user request may trigger another lookup after cache expiry or eviction due to capacity limits, a server restart, processing by another process, or a previous lookup failure.

Collected metadata and images are not saved to the database or upload storage. The original URL and card identifier are saved in the content written by the member.