The Agent Skills Directory

[COMMAND_EXECUTION]: The skill generates Python or TypeScript scripts based on provided job patterns and instructs the user to run them locally (e.g., python evaluate.py) to execute the agent comparisons.
[EXTERNAL_DOWNLOADS]: Core functionality depends on installing the evaluatorq and orq-ai-sdk packages from public registries (PyPI and NPM). These are official resources associated with the skill's vendor.
[DATA_EXFILTRATION]: When the ORQ_API_KEY environment variable is present, the skill automatically transmits agent performance data, query responses, and evaluation scores to the orq.ai platform's Experiment UI.
[PROMPT_INJECTION]: The skill processes external data points which are used as inputs for testing agents. While this introduces an indirect prompt injection surface, the risk is mitigated by instructions to use synthetic datasets and structured evaluation templates.

compare-agents