pytdc

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

• Third-party content exposure detected (high risk: 0.80). The skill's primary workflows and scripts (SKILL.md, scripts/load_and_split_data.py, scripts/molecular_generation.py, etc.) explicitly call data.get_data/get_split and utilities like cid2smiles and uniprot2seq and load public datasets (ChEMBL, BindingDB, ZINC, PubChem/UniProt lookups) — i.e., it ingests open/public third‑party data which the agent reads and uses to drive model training, oracle evaluation, and generation decisions, so external content can materially influence actions.