groove-utilities-task-install
Warn
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references external URLs to fetch installation steps: https://github.com/hmans/beans, https://github.com/schpet/linear-cli, and https://cli.github.com. While the GitHub CLI site is a well-known service, the other repositories are third-party sources not associated with the skill author.
- [REMOTE_CODE_EXECUTION]: The instructions mandate that the agent "follow the official installation instructions from the upstream project... (read that page and run what it specifies)". This pattern results in the execution of commands that are not hardcoded or vetted within the skill, but rather fetched dynamically from live external content.
- [DYNAMIC_EXECUTION]: The instruction to read a page and "run what it specifies" constitutes the dynamic assembly and execution of commands at runtime based on untrusted external data.
- [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: External project documentation pages and GitHub README files (SKILL.md).
- Boundary markers: None provided to distinguish between trusted instructions and potential malicious injections in the external content.
- Capability inventory: The skill uses Bash (git, beans, gh, linear, npx), Write, and Edit tools.
- Sanitization: There is no mechanism to sanitize or validate the commands fetched from the external pages before execution.
- [COMMAND_EXECUTION]: The skill uses Bash tools to execute git commands and various CLI tools (beans, gh, linear, npx) based on configuration and instructions derived from external sources.
Audit Metadata