The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local run.sh scripts using subprocess.run with list-based arguments in eval_runner.py. This is the intended behavior for running behavioral tests and uses secure implementation patterns.
[EXTERNAL_DOWNLOADS]: The run.sh script leverages uv, a trusted dependency manager, to execute the Python environment.
[DATA_EXFILTRATION]: Sourcing of a .env file from the project root in run.sh is performed to provide necessary configuration for tests, which is standard in development workflows.

eval-skills