alterlab-gene-db
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches gene information, biological pathways, and functional annotations from external NCBI API endpoints. This data is then processed and returned to the agent, creating a potential vector for indirect prompt injection if the upstream data source were to contain instructions designed to manipulate the agent's behavior.
- Ingestion points: Data is ingested via the E-utilities and Datasets APIs in
scripts/fetch_gene_data.py,scripts/query_gene.py, andscripts/batch_gene_lookup.py. - Boundary markers: The skill does not use specific delimiters or explicit instructions to help the agent distinguish between data and potential commands within the fetched content.
- Capability inventory: The skill has permissions to read files and perform network operations via Python, providing a capability set that could be exploited if an injection were successful.
- Sanitization: The scripts use standard JSON parsing for the API responses but do not perform sanitization or filtering of the text content for prompt injection patterns before outputting it to the agent.
Audit Metadata