archs4-database

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

• Third-party content exposure detected (high risk: 0.90). The skill's SKILL.md shows explicit, required runtime use of the public ARCHS4 REST API (https://maayanlab.cloud/archs4/api/v1, e.g., /samples/search and /meta/genes/… endpoints) to fetch GEO/SRA-derived sample metadata and gene records (public, user-submitted content) which the agent parses and uses to drive downstream decisions (sample selection, co-expression expansion, gene lists), so untrusted third‑party content can materially alter behavior.