alphagenome-variant-impact-score

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides legitimate bioinformatics functionality for variant impact scoring using the AlphaGenome Atlas. The instructions and scripts are well-documented and do not exhibit any malicious behaviors.
  • [EXTERNAL_DOWNLOADS]: The script scripts/alphagenome_atlas_avi.py downloads a reference GENCODE annotation file from a trusted Google Cloud Storage bucket (storage.googleapis.com). This is a legitimate operation required for the skill's primary function of gene annotation.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external genomic data provided by the user (VCF and tabular files). The processing is restricted to data annotation and does not involve the execution of instructions embedded within the input data, representing a low risk surface for indirect prompt injection.
  • [CREDENTIALS_SAFE]: The skill requires an API key for accessing the AlphaGenome Atlas. It correctly instructs the user to store this key in a .env file and uses the python-dotenv library to load it securely into the environment, following standard security best practices for credential management.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 03:01 PM
Security Audit — agent-trust-hub — alphagenome-variant-impact-score