Oracle ADK Expert

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

• Third-party content exposure detected (high risk: 0.90). The skill's Slackbot integration explicitly reads and executes arbitrary user messages from Slack (see "Slackbot Integration" where user_message = event['text'] and response = agent.execute(user_message)), exposing the agent to untrusted, user-generated third‑party content that could enable indirect prompt injection.