Asta Software Experiment Runner
Pass
Audited by Gen Agent Trust Hub on May 8, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installs the 'asta-plugins' package directly from the official Allen Institute for AI (allenai) GitHub repository. This is a trusted organization and the resource is authentic to the skill's author.
- [COMMAND_EXECUTION]: Executes the 'asta' CLI and directory management commands ('mkdir'). These operations are restricted to the '.asta/experiment/' directory and are scoped to the primary purpose of running computational experiments.
- [PROMPT_INJECTION]: The skill accepts user-provided research tasks and background knowledge which are interpolated into command arguments. This presents an indirect prompt injection surface; however, this is a standard design for such a tool and is considered acceptable given the limited scope of the 'asta' command's experimental functionality.
Audit Metadata