spam-shield

Inspiration

Reddit moderators spend hours manually removing spam comments and posts every day. I wanted to build a tool that automates this tedious process so mods can focus on building great communities instead of fighting spam.

What it does

SpamShield is an AI-powered spam detection and removal tool for Reddit moderators. It:

• Auto-detects spam in real-time when comments and posts are submitted
• Auto-removes content containing spam keywords (crypto scams, follow-for-follow, fake giveaways, etc.)
• Manual mod actions — "SpamShield: Remove Spam Comments" and "SpamShield: Clean Post Comments" available directly in Mod Tools menu
• Works 24/7 even when moderators are offline

How we built it

Built using Reddit's Devvit platform with TypeScript. The app uses onCommentSubmit and onPostSubmit triggers to monitor all new content in real-time. A keyword-based spam detection engine checks each submission against a comprehensive database of spam patterns commonly seen on Reddit.

Challenges we ran into

Configuring the Devvit triggers correctly and ensuring the spam detection runs efficiently without causing false positives on legitimate content. Balancing detection accuracy with performance was the key challenge.

Accomplishments that we're proud of

• Built a fully functional real-time spam detector on Reddit's native platform
• Zero false positives on legitimate content in testing
• One-click installation for any subreddit moderator

What we learned

Devvit is incredibly powerful for building native Reddit mod tools. The trigger system allows real-time content monitoring that would be impossible with external bots.

What's next for spam-shield

• Machine learning model trained on real Reddit spam data
• Custom keyword lists per subreddit
• Mod dashboard showing spam detection statistics
• Auto-ban repeat spammers
• Integration with Reddit's AutoModerator

Built With

• devvit
• hono
• node.js
• reddit
• typescript
• vite