The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes a helper script (scripts/wqo.sh) to download and install the wqo package (version 0.1.1) from the official Python Package Index (PyPI). This installation occurs within a project-local virtual environment (.wqo/venv) rather than globally.\n- [COMMAND_EXECUTION]: The skill interacts with the WorldQuant BRAIN API by executing the wqo CLI tool. The execution environment is secured using umask 077 to ensure that created directories and files have restricted permissions. Arguments are passed to the CLI using standard shell quoting ("$@") to prevent command injection.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources, including alpha identifiers and simulation reports from WorldQuant BRAIN. It implements security best practices by requiring the agent to present a full diagnostic report to the user and wait for explicit approval before proceeding with any state-changing submission operations.\n- [DATA_EXPOSURE]: The skill documentation identifies the sensitive credentials file ~/.brain_credentials.json. It explicitly prohibits the AI agent from reading, writing, or logging the contents of this file, directing users to manage their credentials independently using local text editors and secure file permissions (chmod 600).

wqo-submit