pairwise-ranking
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external data objects, including 'candidate_set' and 'pairwise_judgments', which constitutes a potential ingestion surface for untrusted data. However, the skill does not possess the necessary capabilities (such as network access, file system writes, or shell execution) to act upon malicious instructions embedded in that data.
- Ingestion points: Input contract in SKILL.md requires
candidate_set,pairwise_judgments, andranking_objective. - Boundary markers: The execution protocol does not define specific delimiters or "ignore instructions" wrappers for the ingested data.
- Capability inventory: The skill contains no code, tool invocations, or subprocess executions.
- Sanitization: No explicit sanitization or validation logic is defined for the input parameters.
Audit Metadata