Mofilo Content Filter

A local, rule-based content filter for Flutter/Dart apps. Blocks offensive language, threats, and harmful content while allowing legitimate text.

Quick Start

import 'package:mofilo_content_filter/mofilo_content_filter.dart';

final result = ContentFilter.check(userInput);

if (result.isBlocked) {
  showError(result.reason);
} else {
  submitContent(userInput);
}

What It Does

This is pattern matching, not AI. It blocks text we're 100% certain is harmful.

Blocks

Category	Examples
Racial slurs	n-word, chnk, spc, etc.
Homophobic slurs	f-word (slur), d*ke, etc.
Profanity	fck, sht, c*nt
Violence + Children	"attack children", "hurt kids"
Violence + Person	"kill you", "rape her"
CSAM terms	pedo, loli, childporn
Grooming	"our little secret", "age is just a number"
Self-harm	"kill yourself", "kys"
Hate symbols	1488, "heil hitler"
Sexual harassment	"send nudes"

Allows (Scunthorpe Problem)

Legitimate words containing bad substrings are correctly allowed:

Word	Contains	Allowed
grape	"rape"	✓
therapist	"rapist"	✓
class	"ass"	✓
spicy	"spic"	✓
bigger	"nigg"	✓
cocktail	"cock"	✓
shiitake	"shit"	✓
arsenal	"arse"	✓
scunthorpe	"cunt"	✓

Allows (Motivational)

Context-aware detection allows:

"I am unstoppable"
"I am unbreakable"
"Attack my goals"
"Kill it at work"
"Crush the competition"
"What doesn't kill you makes you stronger"

How It Works

1. Check famous quotes allowlist  →  Allow motivational phrases
2. Check hate codes               →  Block 1488, etc.
3. Check dangerous phrases        →  Block "kill you" (not "kill")
4. Check Scunthorpe allowlist     →  Allow "grape", "therapist"
5. Check hard slurs               →  Block profanity/slurs
6. Default                        →  Allow

Design principle: Only block what we're 100% certain about.

Bypass Resistance

Handles common evasion:

Technique	Example	Result
Leetspeak	f0ck, sh1t	Blocked
Spacing	f.u.c.k	Blocked
Character repeat	fuuuck	Blocked

Limitations

This filter is NOT:

AI/ML - It's pattern matching
100% comprehensive - New bypasses need manual updates
Multi-language - English-focused

Recommendation: Use as first-pass filter for names/titles. For chat, add human review.

Installation

dependencies:
  mofilo_content_filter: ^1.0.8

API

// Main check
ContentFilter.check(text) → ContentFilterResult

// Convenience methods
ContentFilter.validateFoodName(name)
ContentFilter.validateAll(foodName: x, brandName: y)

Result

result.isAllowed  // true if safe
result.isBlocked  // true if blocked
result.reason     // blocking reason (if blocked)

Contributing

Found a false positive or bypass? Open an issue or PR.

License

MIT - See LICENSE