cube-build-content
Pass
Audited by Gen Agent Trust Hub on Aug 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches the Cube CLI installation script from the vendor's official GitHub repository at
https://raw.githubusercontent.com/cube-js/cube/master/install-cli.sh. - [REMOTE_CODE_EXECUTION]: Recommends installing the Cube CLI by piping a remote installation script directly into a shell (
curl ... | sh). This is the standard, documented installation method for the vendor's own tooling. - [COMMAND_EXECUTION]: Relies on the
cubeCLI tool to perform various content management tasks such as creating, updating, and deleting workbooks and reports. - [PROMPT_INJECTION]: The skill contains surfaces for indirect prompt injection where untrusted user input is interpolated into CLI arguments like
--json-queryand--data(Ingestion points: SKILL.md). Boundary markers and explicit sanitization are not provided in the skill instructions. The skill possesses capabilities to modify workspace resources and interact with the Cube API (Capability inventory: SKILL.md). This is an inherent surface for skills that wrapper command-line interfaces.
Audit Metadata