using-backstage-backend-services

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.70). High indirect prompt-injection risk: if the skill’s runtime uses coreServices.urlReader to fetch external URLs, the fetched page/file text is converted to readable strings (e.g., via buffer().toString('utf-8')) and then fed into the agent’s LLM context.