slurm-info-summary
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION] (LOW): The skill executes a local shell script that calls standard SLURM binaries (sinfo, scontrol, sacctmgr) to gather cluster metadata. This is appropriate for its stated purpose but involves command execution.
- [FILE_SYSTEM_ACCESS] (LOW): The skill reads and writes a markdown summary file within its own internal references directory (~/.claude/skills/slurm-info-summary/references/). No sensitive system paths are accessed.
- [INDIRECT_PROMPT_INJECTION] (LOW): The skill processes output from system commands. 1. Ingestion: scripts/gather-slurm-info.sh stdout. 2. Boundaries: Structured markers like === SECTION === are used. 3. Capabilities: Local file write and markdown display. 4. Sanitization: The agent parses data into a fixed template. Severity is LOW as the output is display-only.
Audit Metadata