NYC

bioservices

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • EXTERNAL_DOWNLOADS (SAFE): The skill installs the 'bioservices' package from PyPI. This is a standard and reputable library used in the scientific community for biological data integration.\n- COMMAND_EXECUTION (SAFE): The included script 'pathway_analysis.py' performs local file and directory operations to save analysis results. These actions are restricted to the user-specified output directory and use safe path-handling methods.\n- DATA_EXFILTRATION (SAFE): Network operations are directed exclusively to official bioinformatics service endpoints (such as UniProt, KEGG, and NCBI). These requests are necessary for the skill's stated purpose and do not target unauthorized or suspicious domains.\n- INDIRECT_PROMPT_INJECTION (SAFE): While the skill ingests data from external biological APIs, it is primarily used for data retrieval and structured analysis. There are no patterns suggesting that this data is used to override agent instructions, and the sources (trusted scientific databases) present a negligible risk of adversarial content.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:19 PM