Skills
skills/bagelhole/devops-security-agent-skills/container-registries/Gen Agent Trust Hub

container-registries

Fail

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION] (HIGH): The skill utilizes a download-and-execute pattern that fetches a third-party installer script. Evidence: In SKILL.md, the command sequence wget https://github.com/goharbor/harbor/releases/download/v2.9.0/harbor-online-installer-v2.9.0.tgz followed by tar xzvf ... and ./install.sh executes code from an unverified source. Risk: The 'goharbor' organization is not included in the Trusted External Sources list, making the execution of its scripts high-risk.
  • [EXTERNAL_DOWNLOADS] (HIGH): The skill fetches a binary archive from an external URL that is not within the defined trust scope.
  • [COMMAND_EXECUTION] (LOW): The skill uses multiple subprocess commands (docker, aws, az, gcloud). While these are expected for the skill's purpose, they represent a capability surface that must be monitored.
  • [DATA_EXPOSURE] (INFO): The skill refers to sensitive local configuration files such as ~/.docker/config.json. This is used for legitimate credential helper setup but identifies a path containing sensitive credentials.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 16, 2026, 12:48 PM