Prompt leak

Detect leakage of system or developer prompts in model output.

YAML key: prompt_leak
Direction: output

Detect leakage of system or developer prompts in model output.

Configuration

Prop

Type

defend.config.yaml (fragment)

guards:
  output:
    enabled: true
    provider: claude
    modules:
      - "prompt_leak"

Configure under guards.output.modules with output guarding enabled and provider claude or openai.

Secrets

Detect likely secrets (API keys, tokens) in user input.

Secrets output

Detect secrets in model output.