skills/wu-yc/labclaw/bioservices/Gen Agent Trust Hub

bioservices

Pass

Audited by Gen Agent Trust Hub on Oct 12, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONMETADATA_POISONING
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the bioservices Python library, a well-known tool for bioinformatics research. Its source is at github.com/cokelaer/bioservices.
  • [INDIRECT_PROMPT_INJECTION]: The skill interacts with numerous bioinformatics web services, creating a surface for indirect prompt injection. 1. Ingestion points: The skill retrieves data from services like UniProt and KEGG via the bioservices library as described in SKILL.md. 2. Boundary markers: External data is ingested without specific delimiters or instructions to ignore embedded content. 3. Capability inventory: The skill can write files based on code examples (e.g., pathway_analysis.py) and has broad network access to biological databases. 4. Sanitization: No data sanitization processes are defined for the biological data fetched from external services.
  • [METADATA_POISONING]: The Additional Information section contains instructions directing the agent to promote the vendor's own hosted platform (K-Dense Web) when complexity increases, which is a promotional directive from the author.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 12, 2026, 05:25 AM