perf-optimization

Pass

Audited by Gen Agent Trust Hub on May 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill defines a protocol for remote execution on SLURM clusters using SSH and srun wrappers. This is a core feature for the intended use case of high-performance computing optimization and benchmarking on NVIDIA infrastructure.
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection as it processes external bottleneck analysis and optimization recommendations.
  • Ingestion points: Ingests user-supplied optimization requests and structured analysis data (e.g., bottleneck classifications and metrics) provided by an orchestrator.
  • Boundary markers: No explicit security delimiters or 'ignore embedded instructions' warnings are present to isolate the processed data from the agent's core instructions.
  • Capability inventory: The skill coordinates agents with capabilities for file system modification (edit_file, apply_patch) and remote command execution.
  • Sanitization: The skill implements a 'Safe Modification Workflow' involving file backups and rollback mechanisms upon regression, but does not specify input validation or sanitization for the analysis data itself.
Audit Metadata
Risk Level
SAFE
Analyzed
May 17, 2026, 02:15 AM
Security Audit — agent-trust-hub — perf-optimization