ngs-shotgun-metagenomics

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • Command Execution: The skill is designed to run various local Python scripts and standard bioinformatics tools like Kraken2 and Nextflow. These operations are core to the skill's purpose of processing genomic data.\n- External Resource Utilization: The skill references nf-core/taxprofiler, a pipeline from a widely recognized bioinformatics community. This is a standard method for ensuring reproducibility and adheres to community practices for genomic analysis.\n- Data Ingestion Surface: The workflow ingests external data from user-supplied CSV and TSV files for sample management and metadata. As these files are provided by the user, this represents a data ingestion surface where input validation is a standard consideration for maintainable tool behavior.\n
  • Ingestion points: shotgun_samples.csv and sample_metadata.tsv (referenced in SKILL.md).\n
  • Boundary markers: No specific boundary markers or "ignore embedded instructions" warnings are visible in the provided instructions.\n
  • Capability inventory: The skill executes local Python scripts (run_shotgun_metagenomics.py) and CLI tools (nextflow, kraken2).\n
  • Sanitization: The provided file does not specify sanitization or validation logic, which typically resides within the implementation of the invoked scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 05:33 AM
Security Audit — agent-trust-hub — ngs-shotgun-metagenomics