google-cloud

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow calls Google Cloud REST APIs (e.g., Logging entries:list returning textPayload/jsonPayload, and Secret Manager versions/latest:access returning decoded payload data), which are outsider-authored free text originating from other users/systems in the target project and can be ingested into the agent’s LLM context.