Global Chat: fuzzy detection for common phished brands
Adds brand-name detection that recognizes common look-alike tricks (swapped letters, extra symbols, mixed case) used in phishing. This helps flag suspicious mentions of popular services without blocking ordinary conversation. The brand list can grow over time as new targets emerge.
Changed
Global Chat anti-phishing: correctly detects risky links while allowing normal links
Link scanning has been refined to better read and evaluate links, including lightly disguised ones, while letting everyday links through. When something looks unsafe, the bot explains what happened so users and staff aren’t left guessing.
Global Chat attachments: display from the correct proxy hostname
Images, videos, and other attachments in Global Chat now open reliably under the intended proxy domain for a consistent, working experience.
Message Logs attachments: display from the correct proxy hostname
Attachments recorded in moderation logs now use the proper proxy domain as well, so reviewers can open evidence without dead links or mixed-domain issues.
Global Chat automod case creation
Automod now opens user-level cases only (not guild-level) when it intervenes. This keeps logs tidy and avoids case spam at the guild level while preserving a clear trail tied to the individual user.