weights-and-biases

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly shows using W&B Artifacts and run.use_artifact()/artifact.download() and artifact.add_reference(...) (including HTTP/S3/GCS references and use of other projects/runs/public reports), which downloads and ingests arbitrary external URLs and public project content, so the agent would read untrusted third-party content as part of its workflow.