tooluniverse-rnaseq-deseq2

Fail

Audited by Gen Agent Trust Hub on Mar 8, 2026

Risk Level: HIGHCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: Multiple instances of insecure command generation were detected where user-supplied file paths are interpolated directly into code strings that are later executed.\n
  • In scripts/convert_rds_to_csv.py, the rds_path argument is inserted into an R script template using an f-string: result <- readRDS(\"{rds_path}\"). This script is then executed via subprocess.run(['Rscript', ...]).\n
  • In references/data_loading.md, example code for rpy2 demonstrates the same vulnerability: ro.r(f'df <- as.data.frame(readRDS(\"{rds_file}\"))').\n
  • These patterns allow an attacker to execute arbitrary R or system commands by providing a specially crafted filename (e.g., using quotes and semicolons to break out of the string and run additional commands).\n- [EXTERNAL_DOWNLOADS]: The skill utilizes and suggests the installation of several standard bioinformatics libraries.\n
  • Dependencies include pydeseq2, gseapy, pandas, numpy, scipy, anndata, statsmodels, and tooluniverse.\n
  • These packages are sourced from standard repositories (PyPI) and are well-established within the scientific community.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Mar 8, 2026, 09:10 PM