beautiful-cli-design
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides detailed ed uc ational content and reference guides for building com-mand-line interfaces using mature ec osy stem tools like Ink, Bub ble Tea, and Gum. It em phasizes best practices such as cursor clean up, TTY detection, and color degradation.
- [INDIRECT_PROMPT_INJECTION]: The skill has a confir med attack surface for indirect prompt injection due to its fun ction al ity as a code generation and ed iting as sistant. In gestion Point s: Un trusted data enters the context via Read, Glob, and G rep tools used to analyze user project s. Boun dary Mark er s: The re are no specific instr uc tion s to delimit or warn the agent about un trusted content in read file s. Capability In ven tor y: The skill is eq ui p ped with Bash and Write tool s, enabling sub proc es s ex ec ution and file sy stem modification. Sanitiz ation: The instr uc tion s do not pres cribe validation or es caping of external content before proc es sing.
Audit Metadata