idea-generation

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

• Third-party content exposure detected (high risk: 0.80). The skill's workflow (Step 3: Novelty Assessment) and the included scripts/novelty_check.py explicitly query and ingest results from the public Semantic Scholar API (and suggests manual searches on Semantic Scholar/arXiv), and those returned paper abstracts/lists are used to decide whether an idea is "Novel" or "Not Novel", so untrusted third‑party content is read and can materially influence agent decisions.