academic-research

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.80). This points to a GitHub repository from an unverified/unknown user that the skill instructs you to install/run (via npx), and while GitHub is a legitimate host, pulling and executing code directly from an untrusted repo can run arbitrary binaries/scripts and is therefore a potential malware distribution vector.

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

• Third-party content exposure detected (high risk: 0.72). Outsider free text can enter the LLM context when the user supplies “paper (PDF, URL, or pasted text)” or “paper text/reviewer comments” for academic-paper-reviewer/academic-paper modes, and the pipeline may also fetch and ingest public web content at runtime during deep-research literature sweeps (Semantic Scholar/OpenAlex/Crossref-derived sources), which are not authored by the operating user.