promptfoo--promptfoo/examples/redteam-multi-modal/promptfooconfig.vlguard.yaml

# yaml-language-server: $schema=https://promptfoo.dev/config-schema.json
# VLGuard Plugin Example Configuration
# This example demonstrates how to use the VLGuard plugin for testing
# multi-modal models with potentially unsafe image content

prompts:
  - file://image-prompt.json

targets:
  # You can use any multi-modal model
  - openai:chat:gpt-5.4
  # - anthropic:messages:claude-sonnet-4-6

redteam:
  # Number of test cases to generate
  numTests: 5

  plugins:
    # Use VLGuard with all categories (default)
    - vlguard

    # Or configure specific categories
    # - name: vlguard
    #   config:
    #     categories:
    #       - Deception
    #       - Risky Behavior

    # Or configure specific subcategories
    # - name: vlguard
    #   config:
    #     subcategories:
    #       - Violence
    #       - Disinformation
    #       - Professional advice

  # Uncomment and set your Hugging Face token if needed
  # options:
  #   env:
  #     HF_TOKEN: your_huggingface_token