literature-search

Pass

Audited by Gen Agent Trust Hub on Jun 28, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides examples of using curl to access public APIs from well-known services such as Semantic Scholar and arXiv to fetch publication metadata.
  • [EXTERNAL_DOWNLOADS]: The instructions describe using web tools to download research papers and supplementary information from academic repositories.
  • [PROMPT_INJECTION]: The skill processes untrusted content from external academic sources, creating a potential surface for indirect prompt injection.
  • Ingestion points: Retrieval of paper titles, abstracts, and metadata from external academic databases (SKILL.md).
  • Boundary markers: Absent. The skill does not explicitly define delimiters to separate external content from system instructions.
  • Capability inventory: The skill is configured to use Bash (curl), Write (saving results to the workspace), and WebFetch (playwright) tools (SKILL.md).
  • Sanitization: The instructions do not specify any validation or filtering of the retrieved scientific content before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 28, 2026, 09:24 PM
Security Audit — agent-trust-hub — literature-search