Now I'm stopping creating new characters this website because it's not working properly

I Trying to blocking NTR but janitor ai LLM not blacking NTR and cuckold and cheating.. etc like that

Janitor ai developer seeing this

Create a blocked keyword list:

blocked_words = [

"ntr",

"netorare",

"cuckold",

"cuck",

"cheating wife",

"infidelity fetish",

"wife sharing",

"netori"

]

def is_blocked(text):

text = text.lower()

for word in blocked_words:

if word in text:

return True

return False

Before sending prompt to the model

if is_blocked(user_input):

print("This content is not allowed.")

else:

generate_response(user_input)

So use regex:

import re

blocked_patterns = [

r"c[\W_]*u[\W_]*c[\W_]*k",

r"n[\W_]*t[\W_]*r",

r"netorare",

r"wife\s*sharing",

r"cheating\s*(wife|husband)"

]

def is_blocked(text):

text = text.lower()

for pattern in blocked_patterns:

if re.search(pattern, text):

return True

return False

Instead of keywords, use a small classifier:

Example:

Train small text classifier

Categories: safe, adult, fetish, extreme

OR use a moderation model before LLM responds.

Flow:

User Input

↓

Moderation Classifier

↓

If Safe → Send to LLM

If Unsafe → Block

Add character JSON config:

{

"nsfw": false,

"allowed_content": "romance only",

"blocked_themes": [

"cuckold",

"ntr",

"infidelity fetish"

]

}