Semantic Scholar Lookup

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly fetches and ingests content from the public Semantic Scholar index (e.g., via the "asta papers snippet-search" and "asta papers search/get/citations" commands described in SKILL.md), including snippet.text which returns ~500-word excerpts from paper full text, and the agent is instructed to read and act on those excerpts to find claims/evidence, so untrusted third-party content can directly influence tool use and decisions.