Hushlist

A silent moderation toolkit - shadow ban disruptive users with scoped bans, graduated enforcement, per-post comment hushing, and seamless legacy import - all synced to AutoModerator.

Comment

Inspiration

Banning disruptive users outright often creates more problems than it solves. Alt accounts flood in, modmail becomes a battleground, and the disruption escalates. Even with modmail mute, the whack-a-mole cycle continues.

We needed a quieter approach, one where problematic accounts are silently filtered while moderators retain the ability to monitor and review their activity later. No confrontation, no escalation. Just a reversible, invisible action that buys the mod team time.

The other side of the problem is high-traffic posts. AMAs, breaking news, anticipated movie releases, matches, election threads, these attract volume that can spiral into toxicity fast. Locking or removing the post punishes everyone. We wanted a way to hush just the comments, keeping the post visible while cooling things down.

Hushlist was built to solve these real moderation problems in a real community. Every feature exists because a moderator needed it.

What it does

Hushlist is a comprehensive shadow ban management tool with two core capabilities:

Shadow Banning with Precision:

  • Scoped bans - Full (posts + comments), posts only, or comments only. Not every situation needs a nuclear option.
  • Temporary bans - 7 days, 30 days, 90 days, or custom durations with automatic expiry and cleanup.
  • Graduated enforcement - A strike system that auto-escalates repeat offenders: first offense gets comments-only for 7 days, second gets a full 30-day ban, third strike is permanent. Moderators can always override.
  • Bulk operations - Add multiple users at once, select and remove in batch.

Comment Hushing:

  • Silence all comments on specific posts without locking or removing them.
  • One-tap hush/unhush from the post context menu, or manage via the dashboard.
  • Ideal for ended AMAs, controversial threads, or high-traffic events.

Everything syncs to AutoModerator - Yet, your existing AutoMod config is never touched. Every action is also logged as a Reddit Mod Note for cross-tool visibility.

The full web dashboard provides user management, analytics, activity logs, an AutoMod scanner for importing legacy shadow ban lists, and sync controls, all in one place.

How we built it

Built entirely on Reddit's Developer Platform (Devvit) using the Devvit Web architecture, to work across devices.

Challenges we ran into

  • Working with real moderator data - Real subreddit configurations are messy and inconsistent. Building import and scanning tools that handle edge cases gracefully meant testing against actual community setups, not clean examples.
  • Cross-platform consistency - Getting the dashboard and navigation to behave the same on Android, iOS, and desktop took more work than expected.
  • Balancing automation with moderator control - Features like auto-escalation are powerful, but they can't override a moderator's explicit decision. Finding the right boundary between helpful and intrusive took iteration.

Accomplishments that we're proud of

  • It just works in the background - Once a moderator takes action, enforcement happens without the app needing to stay active.
  • Smart defaults that respect overrides - The graduated enforcement system handles repeat offenders automatically, but any explicit moderator choice takes priority.
  • Seamless migration from existing setups - Moderators can scan their current configuration, preview what would be imported, and bring it into Hushlist without re-entering anything or disrupting what's already in place.

  • Zero-latency enforcement - Content is filtered at Reddit's infrastructure level. No latency, no gaps.

  • Bidirectional sync that doesn't break things - Much harder than it sounds.

What we learned

  • AutoModerator's wiki format has a lot of undocumented edge cases. Building robust parsing for it taught us to test against real configs, not idealized examples.
  • Powerful moderation tools need built-in guardrails. Temporary bans with auto-expiry and graduated enforcement prevent actions from outlasting their purpose. The tool should nudge toward proportionate responses, not just enable maximum force.

What's next for this project

  • Smarter nudges for moderators when enforcement patterns suggest escalation beyond the app
  • More granular controls for subreddits to customize which features are active
  • Deeper visibility into user history within the dashboard

Built With

Share this project:

Updates