lit-review
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the processing of untrusted external data, which creates a surface for indirect prompt injection.
- Ingestion points: The agent fetches and reads full-text documents from external repositories, preprint servers (e.g., arXiv, bioRxiv), and individual author websites as specified in SKILL.md.
- Boundary markers: There are no instructions to use delimiters or ignore embedded prompts when parsing the content of these papers.
- Capability inventory: The skill employs search APIs and document processing tools (e.g., Semantic Scholar, Google Scholar) to extract quotes and synthesize research novelty claims.
- Sanitization: The protocol does not include methods for sanitizing or filtering external text before interpretation by the agent.
Audit Metadata