databricks-mlflow-evaluation

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

• Third-party content exposure detected (low risk: 0.10). The workflow involves searching and analyzing traces from Databricks experiments and Unity Catalog tables, which can contain user-submitted query inputs and chat data, creating a low-risk exposure path where an attacker's free text input could be ingested by the agent during evaluation or trace review.