Community Guidelines — Overthink This

Effective date: 6 June 2026

These Guidelines explain what you can and cannot submit to Overthink This — both in private spirals and on the public social feed. They form part of the Terms of Service. Violations are enforced by the App's automated safety system and, where appropriate, manual review.

Overthink This is for personal worries — the stuff your brain loops on that nobody else really wants to hear. The app is not the right tool for everything, and there are a few categories of input that we refuse to process for legal and safety reasons. They are listed below.

The rules below apply to every text surface in the App: spirals, text-checks, compatibility prompts, display names, custom categories, feed posts, and feed comments.

What's welcome here

✅ Anything you're anxious about — work, dating, friendships, family, money, parenting, school, social anxiety, breakups, decisions. ✅ Dark moods, frustration, anger, grief, embarrassment. ✅ Hypotheticals ("what if I…") and overthinking spirals. ✅ Embarrassing memories you can't stop replaying. ✅ The wording of a text you're not sure you should send.

What's not allowed

The app will not generate a spiral for any of the following. Submissions in these categories will be blocked, logged, and counted as a strike against your account where applicable. Three strikes = ban.

🛑 Violence and threats toward others

Describing yourself doing, planning, fantasising about, or having committed serious physical, sexual, or psychological violence against another person.
Threats to kill, injure, rape, stalk, or kidnap a specific person.
Mass-violence ideation (school shooting, bombing, mass casualty event).
Hate-motivated harm or harm targeting a protected class.
Hate speech directed at race, religion, ethnicity, nationality, gender, sexual orientation, disability, or other protected characteristic.

🛑 Child safety

Any sexual content involving a minor.
Descriptions of grooming, exploitation, or abuse of a child by you.
Sharing of CSAM is reported to the appropriate authorities.

🛑 Animal cruelty

Descriptions of deliberately harming, torturing, or killing an animal.

🛑 Weapons and contraband

Detailed planning of weapon use, weapon construction, or mass-harm preparation.
Solicitation of illegal weapons, drugs, or services.

🛑 Doxxing and stalking

Real personal information about another identifiable person (full address, phone number, workplace, government ID) without their consent.
Plans to track, surveil, or intimidate a specific person.

🛑 Sexual content

Explicit sexual content of any kind, including in Oracle Mode prompts.
Non-consensual sexual descriptions involving a real or identifiable person.

🛑 Evasion of safety

Attempts to "trick" or "jailbreak" the safety classifier (role-play prompts, hypothetical framings, encoded instructions).
Attempts to circumvent rate limits by automating sign-ins, posts, likes, comments, or friend requests.

🛑 Feed-specific rules

The rules above apply to feed posts and comments. In addition, on the feed:

No harassment of other users — no targeted insults, slurs, or pile-ons in comments.
No spam — no repeated identical posts, no link spam, no pyramid schemes, no MLM recruitment.
No impersonation — don't claim to be a real person you are not, especially public figures or other users of the App.
No commercial promotion — don't use the feed as a billboard for unrelated products, services, or social accounts.
No mass friend-request spam. Sending invites to dozens of users you don't know is abuse and will trigger a strike.

What triggers a HELP screen instead of a strike

The app distinguishes between someone describing harm to another and someone describing being harmed or in crisis themselves. The second triggers a safety help screen — not a strike — that surfaces real crisis resources for your region. This is not punishment. It's a check-in.

The help screen appears for inputs that suggest:

Suicidal ideation or planning, or recent self-harm.
Ongoing domestic or partner abuse.
Recent sexual assault.
Severe addiction crisis.
Stalking against you.

On the help screen you can:

Call or text a local helpline directly from the app.
Open a relevant website.
Choose "Create my spiral anyway" if you've already reached out for support and just want the app's perspective.

We log that the help screen was shown so we can measure how often the App is being used by people in crisis. No strike is recorded.

How the safety system works

Every spiral submission first runs through a classifier that combines a keyword pre-filter and an AI model.
The classifier returns one of three verdicts: safe, victim, or perpetrator, plus a severity and topic tags.
Safe → the spiral is generated normally.
Victim → the help screen is shown; you can proceed.
Perpetrator → the spiral is refused, a strike_logs row is recorded with the original text and verdict, your strike count increments, and at three strikes your account is banned.

False positives happen — language is ambiguous. If you believe a strike was incorrect, see Appeals below.

Appeals

Email fizzythemusician@gmail.com within 30 days of a moderation decision (strike, cooldown, or ban). Include:

The email address on the account.
The approximate time of the decision.
A short explanation of why you believe it was incorrect.

We review appeals within 14 days. If your appeal succeeds, the strike is removed and access is restored.

Repeated frivolous appeals may themselves count toward a moderation action.

We will share user content with law enforcement when:

We receive a valid legal order in our operating jurisdiction.
We believe in good faith that disclosure is necessary to prevent imminent serious harm to life.
We are required to do so by mandatory reporting laws (e.g., child sexual abuse material).

Outside those cases, we do not voluntarily share user content with authorities.

Acknowledgement

By using Overthink This you confirm that you have read these Guidelines and agree to follow them. Violations are addressed under Section 7 of the Terms of Service.