Module patterns 

Pattern definitions for LLM security detection

Structs

DANGEROUS_KEYWORDS

Dangerous keywords that should trigger warnings

PROMPT_INJECTION_PATTERNS

Detect prompt injection attempts with DoS protection

RTL_OVERRIDE_CHARS

RTL override characters

SUSPICIOUS_OUTPUT_PATTERNS

Output validation patterns

Functions

get_dangerous_keywords

Get all dangerous keywords

get_prompt_injection_patterns

Get all prompt injection patterns

get_rtl_override_chars

Get RTL override characters

get_suspicious_output_patterns

Get suspicious output patterns