Every link runs the gauntlet before anyone clicks.
One phishing URL. Four checkpoints. AntiLink pushes every message through a live scanning pipeline - normalize, heuristics, verdict, with threat-intel a command away - so a look-alike domain never survives the trip to #general.
Core scanning, the master switch, invite blocking, and external-link controls are Free. Enforcement modes require AntiLink Plus / Premium.
stage 01
Normalize
The raw URL is canonicalized - protocol, host, punycode and path all made comparable.
stage 02
Offline heuristics
Zero-latency scoring runs on every message: homoglyph fold, typosquat distance, hostname entropy, credential-in-URL and risky TLDs.
stage 03
Threat-intel feeds
On an on-demand /checklink lookup, an inconclusive link is cross-checked against live feeds - URLhaus and VirusTotal. Deliberately kept off the message hot path so live scanning stays instant.
stage 04
Verdict
Scores collapse into one level - clean, low, medium, high or critical - recorded on every block. Your policy decides what gets removed, and your enforcement mode decides what happens to the member.
Every message and every edit runs stages 01, 02 and 04 in the background, with no network call in the way. Webhook posts have their own guard, and stage 03 runs on an on-demand /checklink lookup.
Check any link yourself. Get the receipts.
Run /checklink on a URL and AntiLink returns a verdict with the exact reasons behind it. It is a reputation check, never a promise of safety.
Risk level
Critical
Reasons
newmember today at 4:12 PM
claim your free nitro disc0rd-nitro.ru/gift
AntiLink bot
Blocked a malicious link from @newmember. Homoglyph domain flagged critical. Warning 1 / 3 issued.
Threats that hide outside plain text
Attackers get creative. Switch these on and the pipeline catches them too.
Scam Pattern Guard
Flags gift-scam and phishing wording sitting next to a link, even when the URL itself looks ordinary.
QR / Attachment Guard
Decodes QR codes inside images and risk-checks the URL hidden in the pixels.
Webhook Abuse Guard
Catches a compromised webhook firing malicious links into your channels. Switch it on in the Security Center.
You set the rules. AntiLink enforces them.
Tune the whole pipeline from the dashboard - trust the domains you choose, then pick exactly what happens to everything else.
Core toggles
Link & phishing protection
Scan every message for malicious URLs
Block Discord invites
Remove unwanted server invites
Block external links
Restrict outside links to chosen channels
Enforcement mode
Whitelist
Unsaved changes
Turn the pipeline on in seconds
Add AntiLink and every message starts flowing through the scanner immediately. Free to start, with Plus and Premium unlocking enforcement modes and the advanced guards.