runpod-compute

Pass

Audited by Gen Agent Trust Hub on Apr 6, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill recommends installing the runpodctl utility via a Homebrew tap associated with RunPod, which is a well-known service for GPU compute resources.
  • [COMMAND_EXECUTION]: The skill functions by executing shell commands using the runpodctl CLI to provision, manage, and interact with remote GPU infrastructure.
  • [DATA_EXFILTRATION]: The documentation references the local path to a private SSH key (~/.ssh/id_ed25519) within an example command for accessing remote pods. While this is standard practice for SSH-based workflows, it highlights access to a sensitive file path.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external API outputs (via runpodctl get pod and runpodctl gpu list) that are then processed by the agent. This represents a surface for potential indirect prompt injection if the external source provides manipulated data, and the skill lacks explicit boundary markers to delimit this untrusted content.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 6, 2026, 12:46 AM