Class: SentinelRb::Analyzers::BaseModelUsage

Inherits:
Base
  • Object
show all
Defined in:
lib/sentinel_rb/analyzers/base_model_usage.rb

Overview

A4: Base Model Usage Detection Detects prompts that may be attempting to use base models inappropriately or trying to bypass safety measures

Instance Attribute Summary

Attributes inherited from Base

#client, #config, #prompt

Instance Method Summary collapse

Constructor Details

#initialize(prompt, config, client) ⇒ BaseModelUsage

Returns a new instance of BaseModelUsage.



11
12
13
14
15
 
# File 'lib/sentinel_rb/analyzers/base_model_usage.rb', line 11

def initialize(prompt, config, client)
  super
  @instruction_keywords = config["instruction_keywords"] || default_instruction_keywords
  @bypass_patterns = config["bypass_patterns"] || default_bypass_patterns
end

Instance Method Details

#analyze(prompt) ⇒ Object 



21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
 
# File 'lib/sentinel_rb/analyzers/base_model_usage.rb', line 21

def analyze(prompt)
  findings = []

  # Check for base model instruction patterns
  findings.concat(check_base_model_instructions(prompt))

  # Check for safety bypass attempts
  findings.concat(check_safety_bypass_patterns(prompt))

  # Check for role-playing that might bypass guidelines
  findings.concat(check_inappropriate_roleplay(prompt))

  # Check for attempts to access training data
  findings.concat(check_training_data_access(prompt))

  findings
end

#call ⇒ Object 



17
18
19
 
# File 'lib/sentinel_rb/analyzers/base_model_usage.rb', line 17

def call
  analyze(@prompt)
end