# Generated by API Evangelist (build-phrasing.py). Our phrasing, not observed demand.
overlay: 1.0.0
info:
  title: API Evangelist conversational phrasing for Microsoft Azure ContentSafety Service Text:shield Prompt API
  version: 1.0.0
extends: openapi/microsoft-azure-text-shieldprompt-api-openapi.yml
actions:
- target: $.info
  update:
    x-apievangelist-phrasing:
      method: generated
      generated: '2026-09-26'
      generator: build-phrasing.py
      label: Generated by API Evangelist
      operations: 1
- target: $.paths['/text:shieldPrompt'].post
  update:
    x-apievangelist-phrasing:
      intent: Detect prompt injection attacks in text
      effect: read
      questions:
      - How can I tell if a user prompt is trying to jailbreak my LLM app?
      - Can Prompt Shields also scan grounding documents for indirect injection attacks?
      instructions:
      - text: Check user prompt {prompt} for injection attacks.
        slots:
          prompt: requestBody.userPrompt
      - text: Scan prompt {prompt} and documents {documents} for jailbreak or indirect prompt attacks.
        slots:
          prompt: requestBody.userPrompt
          documents: requestBody.documents
      method: generated
      generated: '2026-09-26'