VibeCheckPro

A proactive background Devvit app that monitors subreddit comment velocity, automatically alerting moderators to heating threads to save time.

Comment

Inspiration

A few nights ago, I was mindlessly scrolling through Reddit and watched a perfectly normal discussion thread suddenly turn into an absolute dumpster fire. By the time the moderators finally showed up to lock it hours later, the damage was already done and the comment section was ruined. It hit me: community management on massive platforms is fundamentally reactive. Moderators are flying blind, usually only finding out about toxic threads after users start spamming the report button. I wanted to build a proactive "early warning radar." Instead of waiting for the community to burn down, my goal was to engineer a system that monitors the pulse of a subreddit in real-time, alerting the mod team the exact second a thread's comment velocity spikes, long before the situation spirals out of control.

What it does

VibeCheck Pro is an autonomous background moderation engine operating entirely within the Reddit ecosystem via Devvit. It silently monitors the onCommentSubmit event stream across a subreddit. Using a custom threshold configured by the moderation team, it tracks the exact submission velocity of comments on every single post. If a post suddenly spikes (for example, receiving 10 comments in 60 seconds), the engine bypasses the standard reporting queue and immediately triggers a high-priority Modmail alert (createModInboxConversation). This alert contains deep links and precise metrics, allowing human moderators to intervene, monitor, or lock the thread before the situation escalates into a moderation crisis.

How we built it

The core technical achievement is our high-performance velocity tracker. Instead of querying a heavy database for every comment, we built a highly optimized sliding time-window bucket system using Devvit's internal Redis datastore. By dynamically linking temporary state keys to the specific post and the current clock minute, and enforcing strict 5-minute time-to-live expirations, the engine achieves real-time analytics at scale with zero database bloat.

That completely removes the weird formatting triggers while keeping the exact same engineering flex. Swap that in, check the preview, and you should be good to go!

Challenges we ran into

Shifting from traditional backend development to Devvit's edge-computing paradigm required a mental pivot. We had to ensure our Redis increment logic (redis.get and redis.set) was entirely type-safe and capable of handling high concurrency traffic without dropping counts or crashing during viral spikes. Furthermore, mapping the Hono routing structure to interface perfectly with Devvit's strict custom menu UI definitions required deep-diving into the SDK's shared types and tearing down a lot of boilerplate code to build a truly custom solution.

Accomplishments that we're proud of

Taking this from a blank terminal to a fully deployed, production-ready analytics engine in a matter of hours is our biggest win. Successfully integrating dynamic Redis state tracking with a seamless, native Reddit configuration form and watching the live background trigger fire perfectly into Modmail during our stress tests was an incredible feeling.

What we learned

We gained massive insight into edge computing infrastructure and manipulating Redis for efficient time-series tracking. We also mastered the Devvit SDK's asynchronous trigger system and learned how to build native-feeling UIs that seamlessly write to a global configuration state without requiring server restarts.

What's next for VibeCheckPro

Currently, the engine relies strictly on volume metrics. The ultimate next step is integrating Large Language Models (LLMs) and Machine Learning pipelines to evaluate the context of the velocity. By feeding the rapid comments into a sentiment analysis model, VibeCheck Pro will be able to distinguish between a healthy, viral discussion and a toxic brigade.

Future updates will include autonomous triage actions such as temporarily locking a thread or auto-collapsing comments if the AI detects both high velocity and high toxicity, fully automating the first line of community defense.

Built With

Share this project:

Updates