genome-analyzer
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill retrieves information from external websites (SNPedia, GWAS Catalog, scientific articles) using a subagent to interpret genetic markers found in the user's local files. This creates a surface where malicious or biased content from these external sources could influence the agent's conclusions or recommendations.
- Ingestion points: Step 1 utilizes a subagent with WebSearch to gather data on SNPs (SKILL.md).
- Boundary markers: Absent; the instructions do not specify delimiters or warnings for the agent to ignore instructions that might be embedded in the fetched research data.
- Capability inventory: The skill utilizes
Globfor file discovery andGrepfor reading file content (SKILL.md). - Sanitization: Absent; the research results are directly mapped to genotypes and incorporated into the final report interpretation without validation or escaping.
Audit Metadata