semantic-scholar

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches and ingests third-party content from the public Semantic Scholar API (e.g., graph/v1/paper/search, paper lookups, citations, recommendations) as shown in SKILL.md, and the agent is expected to read abstracts, tldr summaries, citation data and recommendations that can materially influence its decisions and follow-up actions.