referee-sim
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external scientific documents and project files, which creates an attack surface for indirect prompt injection attacks.
- Ingestion points: The agent is instructed to read the scientific document under audit, as well as associated project files and result notes (referenced in sections 4 and 7 of SKILL.md).
- Boundary markers: The instructions do not define specific delimiters or instructions to the agent to disregard commands embedded within the analyzed content.
- Capability inventory: The skill's primary capabilities include reading text files and writing a local markdown verdict file (e.g., referee_sim_.md). The skill does not appear to use network tools or execute arbitrary shell commands.
- Sanitization: No explicit sanitization, filtering, or escaping of the input document content is described in the procedure.
Audit Metadata