bioservices

Pass

Audited by Gen Agent Trust Hub on Oct 24, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates data retrieval from reputable scientific organizations including UniProt (rest.uniprot.org), KEGG (rest.kegg.jp), and EMBL-EBI (ebi.ac.uk). These are well-known services in the bioinformatics domain.
  • [COMMAND_EXECUTION]: Instructions specify the installation of the 'bioservices' package using the 'uv' package manager. This is standard environment configuration for scientific agent tasks.
  • [INDIRECT_PROMPT_INJECTION]: The skill's attack surface for processing untrusted external data was evaluated.
  • Ingestion points: Processes JSON, TSV, and XML responses from various biological data providers within Python scripts (e.g., scripts/protein_analysis_workflow.py).
  • Boundary markers: None; however, the skill primarily uses structured data parsing (JSON/CSV) which programmatically separates data from instruction context.
  • Capability inventory: Includes file system access (Write, Edit) and shell execution (Bash) for running bundled scripts and managing local analysis results.
  • Sanitization: Scripts utilize standard data loaders (json, csv) and URL encoding (urllib.parse.quote) for outbound requests.
  • [CREDENTIALS_SAFE]: No hardcoded API keys or credentials were found. The skill correctly utilizes an environment variable (NCBI_EMAIL) for user-provided contact information required by the EMBL-EBI BLAST service.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 24, 2026, 09:46 PM