Grizzly-mod-ai

Reddit Developer app listing for grizzly-mod-ai, showing the uploaded Devvit app and feature README.
Queue Whisperer custom post dashboard showing auto-prioritization, queue stats, average risk, and mock fallback data.
Devvit app running in the public test subreddit, with Queue Whisperer posts created for moderator testing.
Prioritized mod queue view ranking high-risk posts and comments with clear “Flagged because” explanations
Moderator context panel with risk score, content preview, quick actions, and detailed risk explanation.
Rule violation history, recent user activity, and moderation action log for transparent review tracking.

Inspiration

Moderators often face long queues where spam, harassment, and repeat offenders are mixed with low-risk reports. Queue Whisperer was built to help mods decide what needs attention first without hiding the reasoning.

What it does

Queue Whisperer prioritizes mod queue items using an explainable 0-10 risk score. It highlights toxic keywords, repeat spam patterns, new-account signals, reports, and recent removal history. Mods get a context panel with user activity, rule violations, risk explanation, and quick actions like approve, remove, warn, and ban.

How we built it

The app is built with Devvit Web, React, TypeScript, Hono, Reddit APIs, and Redis. The server scores queue items and caches moderation settings/action logs. The frontend gives moderators a clean dashboard with color-coded risk levels and an auto-prioritize toggle.

Challenges we ran into

The hardest parts were adapting the app to the latest Devvit Web project structure, making the custom post work inside Reddit, and ensuring the app could still demo safely with mock queue data when live queue data is unavailable.

Accomplishments that we're proud of

The app is uploaded to Reddit Developers, installed in a public test subreddit, and runs as a Devvit custom post. It gives moderators a practical triage workflow instead of just another report list.

What's next for Grizzly-mod-ai

Next steps include subreddit-specific scoring rules, richer historical mod notes, better toxicity providers, and team-level analytics showing how much queue review time the app saves.