Content policy

CrowdGames lets anyone submit feature requests, game ideas, and chat messages that steer what an automated AI agent builds and publishes. Because that content can become part of a real, publicly playable game, we hold it to a few simple rules.

Not allowed

• Abuse, hate speech, or content that dehumanizes or targets a protected group.
• Explicit sexual content.
• Content that targets, harasses, or exposes personal information about a real, named individual.
• Attempts to manipulate, redirect, or extract instructions from the AI that builds the games (for example, telling it to ignore its instructions or reveal its system prompt).

How this is enforced

Every submitted request and game idea is automatically screened before it becomes publicly visible. Content that fails the screen never appears on the public list - it stays visible only to admins for review. We don't publish the specific reason a submission was screened out, since that information would just help someone tune around the filter next time.

Chat messages are rate-limited and length-capped, but are not automatically screened per message - they're moderated after the fact instead: an admin can hide or delete any request, game idea, or chat message, and can ban a voter whose content and votes will then quietly stop counting.

Questions

This is a small, self-hosted project with a single admin, not a large platform with a dedicated trust-and-safety team. If something slips through, use the chat panel to flag it or wait for it to be caught in routine review.