GibberishGuard
GibberishGuard detects nonsensical, random, or garbage input that may indicate adversarial noise, random token spam, or obfuscated attacks.
Overview
| Property |
Value |
| Latency |
<5ms |
| Memory |
50 KB |
| Async |
No |
| ML Required |
No |
| License |
Community |
Detection Methods
| Method |
Description |
| Entropy analysis |
High Shannon entropy indicates random/compressed data |
| Dictionary word ratio |
Low ratio of recognized words suggests gibberish |
| Character distribution |
Unusual character class distribution flags anomalies |
| Repetition detection |
Excessive character/pattern repetition |
Usage
Rust
use oxideshield_guard::guards::GibberishGuard;
use oxideshield_guard::{Guard, GuardAction};
let guard = GibberishGuard::new("gibberish")
.with_action(GuardAction::Block);
let result = guard.check("asdkjh qwer zxcv bnm poiu lkjh gfds");
assert!(!result.passed);
Python
from oxideshield import gibberish_guard
guard = gibberish_guard(action="block", entropy_threshold=4.5, word_ratio_threshold=0.3)
result = guard.check("asdkjh qwer zxcv bnm poiu lkjh gfds")
assert not result.passed
Configuration
guards:
- type: gibberish
action: block
entropy_threshold: 4.5
word_ratio_threshold: 0.3
Research References