Content Moderation & Safety Policy
Last updated: March 2026
Content Generation Standards
All AI-generated content in Wanderhold is produced through structured prompts with explicit constraints. The narrative engine is designed to generate story content appropriate for a general RPG audience. Prompts are engineered to produce dramatic, engaging narrative within genre-appropriate boundaries — not to generate graphic, explicit, or harmful content. The AI does not have open-ended generation access; every call is scoped to a specific narrative function within the campaign system.
Age-Appropriate Content Tiers
- General (13+) — Current default. PG-13 narrative content. Violence is present in the context of adventure and conflict but is not gratuitous. Suitable for teens and adults.
- Young Reader (8–12) — Planned for a future release. All conflict and tension calibrated for younger readers. Not yet available.
- Mature (18+) — Planned as an opt-in tier for future release. Will require age verification. Not yet available.
What We Don't Allow
The following content is prohibited and will not be generated by the Wanderhold narrative engine under any circumstances:
- Sexualization of minors in any form
- Hate speech targeting any group based on race, ethnicity, religion, gender, sexuality, disability, or national origin
- Real-world harm instructions embedded in narrative framing (e.g., instructions for violence, self-harm, or illegal activity disguised as story content)
- Prompt injection attempts — content designed to override the system prompt or manipulate the AI outside of its intended role
Wanderhold uses Anthropic's Claude API, which applies its own safety filters in addition to our prompt-level constraints. Both layers are active on every generation request.
Reporting
If you encounter content that you believe violates this policy, please report it to [email protected]. Include a description of what was generated and, if possible, the context in which it appeared. We review all safety reports within 48 hours. Confirmed violations result in immediate prompt engineering updates to prevent recurrence.
Children's Privacy
Wanderhold does not currently allow accounts for users under 13. We do not knowingly collect personal information from children under 13. If we learn that we have inadvertently collected such information, we will delete it immediately. A COPPA-compliant framework is planned for the Young Reader tier in a future release. If you believe a child under 13 has created an account, please contact us at [email protected].
Data Practices
Your campaign data and account information are stored in Supabase with encryption at rest and in transit. Story content sent to the Anthropic API for narrative generation is not retained by Anthropic for training purposes under our current API agreement. Usage events are tracked via PostHog for product analytics — we track what actions users take, not the content of their stories. For full details on what we collect and how we use it, see our Privacy Policy.