run-cloud-sandboxes

Pass

Audited by Gen Agent Trust Hub on Jul 28, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONCREDENTIALS_UNSAFEDATA_EXFILTRATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the runcloud CLI and @run-cloud/sdk from the NPM registry to interact with the api.run.cloud service. These represent the official tools associated with the skill's vendor and purpose.
  • [COMMAND_EXECUTION]: Instructions describe how to use shell commands and SDK methods to manage sandboxes, including creation, execution of commands inside the VM, and resource cleanup.
  • [CREDENTIALS_UNSAFE]: Authentication is managed via RUN_CLOUD_API_KEY. The skill provides explicit safety instructions, warning the agent to never hardcode, print, or commit credentials and recommending the use of environment variables or interactive login.
  • [DATA_EXFILTRATION]: The skill provides capabilities for file transfers (readFile, writeFile) and network tunneling (openTunnel). These are documented with security warnings to treat URLs as bearer secrets and to revoke access immediately after use.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 28, 2026, 05:36 PM
Security Audit — agent-trust-hub — run-cloud-sandboxes