slurm-job-script-generator

Pass

Audited by Gen Agent Trust Hub on Oct 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied resource requirements and commands to generate .sbatch shell scripts, creating a surface for indirect instructions to be embedded in job parameters.\n
  • Ingestion points: The scripts/slurm_script_generator.py script accepts user input via CLI arguments for job names, partitions, modules, environment variables, and run commands (after the -- separator).\n
  • Boundary markers: The generated script includes #SBATCH directives and standard shell headers, but does not use explicit delimiters to isolate user-provided command strings from the surrounding script logic beyond standard shell quoting.\n
  • Capability inventory: The skill utilizes the Write tool to save the generated shell script to a file path and the Bash tool to execute the generator script. The resulting script is intended for execution on HPC cluster resources.\n
  • Sanitization: The generator implementation in scripts/slurm_script_generator.py applies strict regular expression validation to most input fields (e.g., job names, walltimes, partitions, modules) and employs shlex.quote to escape shell metacharacters in commands and environment variables, significantly reducing the risk of successful injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 22, 2026, 07:59 PM