name: Detect spam
model: openai/gpt-4o-mini
messages:
  - role: system
    content: "" # Since it's not a fix value, it should be generated and replaced at runtime
  - role: user
    content: "" # This will be replaced at runtime