modelopt-launching-evals

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

• Third-party content exposure detected (low risk: 0.10). The required workflow reads benchmark artifacts (tasks.jsonl, output.jsonl, prompt.txt, client/server logs) that carry benchmark task text such as SWE-bench problem statements, which originate from outsider-authored GitHub issues, but only for datasets the user selected in the config, so an attacker can publish poison without forcing delivery.