How it works
Effective community moderation starts with a keyword blocklist — words and phrases that automatically send content to the moderation queue or trigger an instant hold before it appears publicly. The Community Moderation Keyword List Builder helps you create platform-specific keyword lists (for Discord AutoMod, Twitch AutoMod, YouTube Community, Facebook Groups, or Instagram comment filters) with built-in starter lists by community type.
Keyword categories provided: - Spam patterns: "DM me", "free followers", "check my bio", "100% guaranteed", emoji-only patterns, repeated character strings, and URL formats commonly used by bots - Harassment terms: slurs, personal attack phrases, common troll openers — categorised by severity (soft-block vs. immediate ban) - Scam indicators: "send 0.1 ETH", "I just made $[X]", "verified giveaway", financial offer patterns - Off-topic promotion: competitor brand names, unrelated product categories, recruiting language - Self-harm and crisis language: a separate list for support communities following safe messaging guidelines
How to use: 1. Select your platform and community type. 2. Browse the starter list — check or uncheck terms to include in your export. 3. Add custom keywords specific to your community's problem patterns. 4. Export as a plain text list (one keyword per line) for import into your platform's moderation settings.
Platform import formats: - Discord AutoMod: comma-separated list or paste into server settings - YouTube: Community, Blocked Words (one per line) - Facebook Groups: Admin Tools, Keyword Alerts (one per line)
Privacy: all list building runs in the browser. No data is transmitted.
Frequently Asked Questions
- 50–200 keywords is the practical range for most communities. Fewer than 50 misses common spam and harassment patterns. More than 500 starts generating false positives — legitimate posts flagged incorrectly by unintended keyword matches. Review the list quarterly: remove keywords that are generating false positives, add new terms that have appeared in actual violations. A blocklist that's never reviewed gradually becomes either too restrictive or too permissive as language and spam patterns evolve.
- A keyword blocklist sends matching posts to the moderation queue (or auto-removes them) — the poster is aware their post was moderated, and mods review and either approve or remove. A shadowban (or 'ghost ban') hides content from the community without notifying the poster — they can still see their own post, but others cannot. Shadowbans are more controversial and harder to defend transparently. Most community management best practices recommend visible moderation (blocklist + hold queue) over invisible shadowbanning.
- Generally no. Publishing the exact blocklist allows spammers and bad actors to craft messages that evade detection by avoiding specific words. Instead, publish the categories of restricted content (no promotional links, no competitor mentions, no financial solicitation) without listing the exact terms. The rules describe the behaviour; the keyword list enforces it. Members who accidentally trigger the filter and have legitimate posts held will typically reach out to mods for clarification.
- Configure your moderation system to hold flagged posts for admin review rather than auto-delete (at least initially). Review held posts daily and approve legitimate ones. When a legitimate post is incorrectly flagged more than twice by the same keyword, consider removing or narrowing that keyword (e.g., replacing 'buy' with 'buy now', which is more specific to spam). Keep a moderation log — if a member complains, you can show them when their post was held and when it was approved.