runpodctl

Pass

Audited by Gen Agent Trust Hub on Aug 21, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references the official installer for the runpodctl CLI via curl -sSL https://cli.runpod.net | bash in SKILL.md and reference/install.md. It also provides links to download platform-specific binaries from the vendor's official GitHub releases repository. These resources originate from the vendor's own infrastructure and established public registries.
  • [COMMAND_EXECUTION]: The skill is designed to guide the agent in executing shell commands using the runpodctl binary. This includes managing pods, serverless endpoints, and network volumes. The execution is scoped to the intended management capabilities of the tool.
  • [DATA_EXFILTRATION]: The skill documents the runpodctl send and runpodctl receive commands for file transfer, as well as runpodctl model add for uploading model artifacts. These are core functionalities of the workload management tool used for data movement between local environments and remote GPU pods/storage.
  • [REMOTE_CODE_EXECUTION]: The runpodctl update command is documented as a way to keep the CLI tool current. This involves downloading and replacing the active binary from the vendor's release server, which is standard behavior for developer tools of this nature.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 21, 2026, 01:32 PM
Security Audit — agent-trust-hub — runpodctl