​
Available Components

OpenAI Moderation Node

Leverages OpenAI’s content moderation API to analyze and filter potentially harmful content

Simple Prompt Moderation Node

Customizable content filtering based on user-defined rules and denied phrases
Key Features:
  • User-defined deny lists
  • Optional LLM-based similarity detection
  • Customizable error messages
  • Flexible rule configuration
Use Cases:
  • Preventing prompt injection attempts
  • Filtering specific keywords or phrases
  • Implementing custom content policies

​
Benefits

  1. Enhanced Safety
    • Protect against harmful or inappropriate content
    • Prevent misuse of AI systems
    • Maintain ethical guidelines
  2. Policy Compliance
    • Ensure adherence to API provider policies
    • Meet regulatory requirements
    • Maintain consistent content standards
  3. Customizable Control
    • Define specific moderation rules
    • Adapt to project requirements
    • Balance safety with functionality
  4. Seamless Integration
    • Works within existing AI workflows
    • Compatible with various language models
    • Easy to implement and maintain

​
Implementation Guide

  1. Choose Your Moderation Approach:
    • Use OpenAI Moderation for comprehensive content analysis
    • Use Simple Prompt Moderation for custom rule-based filtering
    • Combine both for layered protection
  2. Configure Parameters:
    • Set up appropriate credentials
    • Define custom error messages
    • Specify denied phrases or rules
  3. Integration Points:
    • Add moderation checks before processing user input
    • Implement content filtering before model responses
    • Monitor and log moderation results
  4. Best Practices:
    • Regularly update moderation rules
    • Monitor false positives/negatives
    • Maintain clear documentation of policies
    • Test moderation effectiveness regularly