literature-review
Pass
Audited by Gen Agent Trust Hub on Apr 23, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill consists entirely of markdown instructions and references to external prompt templates. It does not include any executable scripts, shell commands, or configuration files that perform automated actions.\n- [INDIRECT_PROMPT_INJECTION]: The skill identifies an attack surface by ingesting external academic papers, though the risk is minimal in the context of research synthesis.\n
- Ingestion points: External research papers and metadata retrieved by the researcher agent.\n
- Boundary markers: None explicitly defined in this file.\n
- Capability inventory: Writing synthesis reports and provenance logs to the outputs/ directory.\n
- Sanitization: No specific sanitization or filtering logic is documented in this file.
Audit Metadata