citation-verifier
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed for the legitimate purpose of verifying academic citations and detecting hallucinated references.
- [EXTERNAL_DOWNLOADS]: The skill fetches metadata from well-known academic services including CrossRef (api.crossref.org), arXiv (arxiv.org), PubMed (ncbi.nlm.nih.gov), and Open Library (openlibrary.org). These are trusted services used for their intended purpose of bibliographic verification.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted document content (Markdown, LaTeX, etc.) provided by the user. While this represents a potential attack surface, the risk is minimal as the skill uses specific regular expressions to extract structured identifiers rather than executing code or performing sensitive operations based on the document text.
- [COMMAND_EXECUTION]: No suspicious command execution or shell injection patterns were found. The tools used (Grep, Glob, Read) are standard for file analysis within the agent's scoped environment.
Audit Metadata