deeplake-managed

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

• Third-party content exposure detected (high risk: 0.80). The skill's documented workflows explicitly ingest public, user-contributed datasets (e.g., client.ingest(..., {"_huggingface": "mnist"}) in SKILL.md and LeRobot episodes that pull via git-lfs in examples.md) and then use that content for indexing, querying, and training, meaning untrusted third‑party content is fetched and subsequently read/used in decision-making and model/training workflows.