0.7.3

What's New

Sentinel 0.7.3 — Hawk

Sentinel 0.7.3 strengthens protection against Discord scam images — fake Nitro gifts, phishing screenshots, and other visual spam. Every protected server benefits from a platform-wide blocklist of known scams. When a new image slips past that list, Sentinel's built-in image intelligence can score it and route uncertain cases to your mods for a quick Spam / Not spam review so protection improves over time.

Turn on Spam Image Blocklist in your server's AutoMod settings in the dashboard to opt in. No extra setup on your side.

🎉 New Features

Global spam image blocklist — Sentinel maintains a shared database of confirmed scam images. When enabled for your server, matching images are stopped immediately — including near-duplicates that scammers reuse after recompressing or lightly editing the same graphic.

AI-assisted image screening — After the blocklist check, Sentinel can analyze attachments with machine learning trained on real Discord scam visuals. High-confidence hits follow your configured AutoMod action; borderline scores go to mod review instead of guessing.

Mod review queue — Uncertain images post to your mod-log channel with Spam and Not spam buttons. Your team's labels help Sentinel learn what matters in live communities. Member reports need agreement from two moderators before resolving, to reduce noise from false reports.

Report spam image — Members and moderators can right-click a message and report a suspicious image. Reports are rate-limited and feed the same review flow.

Large image support — Boosted Discord servers allow attachments larger than 8 MB; Sentinel scans those for blocklist matches and AI screening.

Deferred checks when AI is busy — If image analysis is temporarily unavailable, Sentinel queues the attachment and rechecks when service is restored, with a notice in mod-log so nothing silently slips through.

🐛 Bug Fixes

None reported from production.

🔧 Improvements

None beyond the new features above.

📋 Summary

0.7.3 Hawk is Sentinel's scam image release: a shared blocklist, AI-assisted screening, mod and member reporting, and support for larger attachments. Open the dashboard → your server → AutoMod and enable Spam Image Blocklist to get started.