The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses standard shell commands mkdir -p and cp to scaffold new test directories. These operations are limited to the tests/functional_tests/test_cases/ directory and are consistent with the intended purpose of automating test setup.
[DATA_EXPOSURE]: The skill reads project-specific configuration files (model_config.yaml, gpt.yaml, moe.yaml). No access to sensitive system files, environment variables, or credentials was detected.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection because it ingests data from external YAML files and uses that data to determine directory names and file content. While no boundary markers or sanitization logic are explicitly defined, the risk is considered low given the skill's specific context within a development environment and the nature of the data being processed.

onboard-gb200-1node-tests