Skills
Navigate
skills/eturkes/claude-scientific-skills/get-available-resources/Gen Agent Trust Hub

get-available-resources

Pass

Audited by Gen Agent Trust Hub on Sep 25, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/detect_resources.py executes local system-level commands to detect hardware capabilities. Evidence includes:
  • Calls to nvidia-smi for NVIDIA GPU detection.
  • Calls to rocm-smi for AMD GPU detection.
  • Calls to sysctl and system_profiler for Apple Silicon detection on macOS.
  • [INDIRECT_PROMPT_INJECTION]: The skill creates a local .claude_resources.json file which the instructions encourage the agent to read to inform its future computational decisions.
  • Ingestion points: Hardware metadata retrieved via psutil and subprocess in scripts/detect_resources.py.
  • Boundary markers: None; the agent is directed to parse the structured JSON directly.
  • Capability inventory: The skill has file-writing and local command-execution capabilities across its scripts.
  • Sanitization: Not explicitly present as the input source is the local operating system diagnostics.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 25, 2026, 02:34 PM