terminal-ui-engineer

Warn

Audited by Gen Agent Trust Hub on Mar 16, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill is designed to generate shell scripts that automatically attempt to install the gum utility using system package managers like apt, pacman, brew, scoop, and winget if the dependency is not found.
  • [COMMAND_EXECUTION]: The generated scripts are explicitly instructed to use sudo for installation on Unix-like systems, which executes commands with elevated administrative privileges.
  • [EXTERNAL_DOWNLOADS]: The skill references and directs the agent to install the gum utility from Charmbracelet's official GitHub repository.
  • [PROMPT_INJECTION]: The skill ingests user preferences and functional goals to define the logic of the generated scripts, creating a surface for indirect prompt injection.
  • Ingestion points: User input provided via the question tool to define the functional goals and actions of the generated script.
  • Boundary markers: None identified; there are no instructions to sanitize input or wrap it in delimiters to prevent the execution of malicious instructions embedded in the goal description.
  • Capability inventory: The generated scripts have the capability to execute arbitrary shell commands, manage system environment variables, and perform file system operations (e.g., piping output to xargs rm).
  • Sanitization: Absent; the instructions do not require the agent to validate or escape user input before incorporating it into the final executable script.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 16, 2026, 04:36 AM