Content Policy

Source review

• Each answer and its reveal context must be supported by a credible, human-verifiable HTTPS source.
• Primary, authoritative, or professionally credible references are preferred; unsourced trivia and social posts are not sufficient.
• Agents may assist with research and review, but they are not accepted as factual sources; every approval must retain explicit, human-verifiable evidence.

Safety review

• Playable questions must be suitable for a family-friendly audience aged 13 and over.
• The deck excludes sexual content, graphic violence, hate, dangerous instructions, and ridicule aimed at a group.
• Historical medicine and obsolete science must be identified as historical, while psychology claims must avoid unsupported pop science.

Language and playability review

• A native-language reviewer checks the wording and pronunciation before publication.
• Prompts should be unfamiliar enough to invite plausible bluffs, while answers must remain clear, checkable, and understandable at reveal time.

Answers stay behind the server boundary

Approved production questions and answers are compiled into the server-only Firebase Functions bundle. The public web build receives locale availability counts, not the production answer bank.

During a round, the backend returns the answer only in the private view of the player assigned to know it. The answer and source become public to the room only when the game reaches the reveal phase.