Mia · Guardrails
Write your limits. Mia never crosses them.
Platform rules are built in and kept current by us. On top, you write your own limits as plain sentences, the way you would brief a new hire. Every reply is checked against all of them before it reaches a fan.
- Plain sentences
- Checked before send
- Any language
My rules, in my words
What you can lock down
Three layers. All checked on every reply.
Your rules
Plain sentences, not a short word list.
A word list catches a word. A rule catches a meaning. "Never say where I live" covers your street, your suburb, your gym and the beach near your flat, in any language, even if nobody typed the exact word. Write rules for the whole workspace once, then add the ones each creator needs. No small cap on how many.
- Workspace rules. Apply to every creator you manage.
- Account rules. Add the extras one creator needs.
- Banned words too. Pet names, places, names, platforms.
- Copy as a template. Same rules across a whole tier.
Platform rules
The prohibited topics, handled for you.
Every reply is checked against the topics adult platforms prohibit, and we keep that list current. We lean toward caution: stopping a borderline line costs far less than letting a real breach reach a fan. Anything that touches a minor ends the chat at once and alerts your team. That is a stop, not a setting.
- Meeting in person, off platform contact or payment.
- Drugs, violence, hate, self harm.
- Non consent, incest or family roleplay, bestiality.
- Bodily waste, public sex, hypnosis.
- Minors. Hard stop, alert, no exceptions.
Off limits, by choice
Things that are allowed, but not you.
Platform rules are the floor, not the whole job. Ban the subjects that would break the spell on an account whose fans know you well: your real city, your ex, a pet name you hate, other creators, other platforms, a topic that is sensitive where your fans live.
Checked before it leaves
Every reply, every time.
Before a reply goes out, Mia checks it: your rules, platform rules, no AI tell, no leaked details, no resent content, a price inside your range. If a draft fails, she rewrites it or stops and hands off. It never reaches the fan first and gets fixed later.
When a fan pushes
She steps out and tells you.
When a fan pushes for something you ruled out, Mia steps out in character, never with a robotic refusal. Then she keeps chatting, closes the chat kindly, or hands off with a brief, depending on the rule. The live wall logs it and the mobile app alerts the roles you picked.
Asked to meet in person. Mia declined in character. Handed to Priya.
Asked for her address. Rule 1. Chat continues.
Problems solved
Compliance, before the fact.
One wrong word can flag an account
Checked before send. Every reply passes platform rules and your rules first, so a breach never reaches the fan.
Fans push for banned roleplay
She declines and hands off. In character, kindly, then your team decides what happens next.
Known creators cannot sound off brand
Rules per persona. Ban the names, places and subjects that would give it away.
Rules drift across many accounts
One workspace list. Write it once, apply it everywhere, add extras per creator.
How it works
Set guardrails in four steps.
Start with the baseline
Platform rules are already on for every account.
Add workspace rules
Write the sentences every creator should follow.
Add account rules
The things one creator never wants said, in her words.
Review the stops
Read stops on the live wall. Mark false alarms and misses. The rules get sharper.
Compliance stops being something you check after the fact.
Why it is better
Rules that understand meaning.
| FanMiind | Typical AI chatter | |
|---|---|---|
| How you write limits | Plain sentences plus words | Word lists |
| Cap on custom limits | No small cap | Often 10 to 30 words |
| Catches the meaning, not just the word | Yes, any language | Exact matches |
| Checks for leaks and AI tells | Yes | No |
| Hard stop on any sign of a minor | Yes | Topic flag |
FAQ
Questions, answered.
Where do I write guardrails?
In the Guardrails tab for each account, or once for your whole workspace. Workspace rules apply to every creator. Account rules add to them for that creator only.
Is there a limit on how many rules I can add?
No small cap. Write the words, phrases and rules each account needs. Most accounts need a handful of sentences, not a long list.
What is blocked by default?
The topics adult platforms prohibit, kept up to date by us: meeting in person, drugs, violence and hate, non consent, incest or family roleplay, bestiality, bodily waste, self harm, public sex, minors, and contact or payment off the platform. Any sign of a minor ends the chat at once and alerts your team.
Why plain sentences instead of a word list?
A word list catches a word, not a meaning. "Never say where I live" covers your street, your suburb, your gym and the beach near your flat, in any language. Mia checks the meaning of every reply, so a rule holds even when nobody used the exact word.
What happens when a fan pushes past a limit?
Mia steps out of the topic in character, then hands off or keeps chatting, as you choose for that rule. The chat moves to the top of the inbox and the mobile app alerts the roles you picked.
Can I block things that are not against platform rules?
Yes. Ban pet names you never use, your real city, your ex, other creators' names, other platforms, topics that are off brand, or subjects that are sensitive where your fans live.
A harmless message was stopped. What now?
Open it from the live wall and tap Not a problem. Mia sends it or rewrites it, and the rule learns. Report anything that slipped through the same way.
Can I copy one account's guardrails to others?
Yes. Set up one account, then copy its rules as a template to every account in the same tier.
Start today
Free for 14 days. Live in two minutes.
Connect an account and watch the AI chatter answer its first fan before your coffee cools.
Start freeNo credit card required.
Book a demo
Let us look at your accounts with you.
We show you where the money leaks today and what the engine would do about it.
Our experts will not bite.