setup
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script located at
{skill_path}/scripts/setup_experiment.pyto handle experiment configuration. It passes user-defined arguments, including an evaluation command string, to this script. - [INDIRECT_PROMPT_INJECTION]: The skill ingest untrusted data in the form of user-provided commands and file paths which are then used in shell execution context.
- Ingestion points: Interactive prompts for 'Domain', 'Target file', and 'Eval command' in SKILL.md.
- Boundary markers: None present in the instructions.
- Capability inventory: Execution of subprocesses via the
setup_experiment.pyscript (SKILL.md). - Sanitization: No explicit sanitization or validation of the user-provided
eval_cmdis defined in the skill instructions.
Audit Metadata