arxiv-reader

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

• Third-party content exposure detected (high risk: 0.90). The skill fetches public arXiv content (RSS/API and raw LaTeX) via arxiv_fetcher.fetcher.fetch_latex_source and _fetch_via_rss/_fetch_via_api, then passes the parsed paper text into the ReaderAgent which feeds it into the LLM chain (agents/reader_agent.py and main.py), so untrusted third‑party paper content is directly ingested and can materially influence the agent's behavior.