The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of Bash and Python scripts to execute Ansible commands. Specifically, scripts/validate_playbook.sh, scripts/validate_role.sh, and scripts/test_role.sh invoke ansible-playbook, ansible-lint, and molecule respectively to process user-provided Ansible files.
[CREDENTIALS_UNSAFE]: Hardcoded credentials are present in the test suite. test/playbooks/bad-playbook.yml contains a password string (hardcoded_password_123), and test/roles/geerlingguy.mysql/defaults/main.yml includes default administrative credentials. These are used as test fixtures to verify that the validator's scanning tools (such as scripts/scan_secrets.sh) correctly identify insecure patterns.
[EXTERNAL_DOWNLOADS]: Several scripts (e.g., scripts/setup_tools.sh and scripts/test_role.sh) automate the installation of validation tools from the Python Package Index (PyPI) and Ansible Galaxy. This includes packages like ansible-core, ansible-lint, checkov, and molecule. These downloads are used to bootstrap the skill's environment.
[REMOTE_CODE_EXECUTION]: The scripts/test_role.sh and scripts/validate_playbook.sh scripts create temporary Python virtual environments and execute pip install to load required libraries at runtime. This behavior is documented and intended for environment readiness.
[SAFE]: The skill includes a copy of the well-known and trusted geerlingguy.mysql Ansible role for testing purposes. All external repository references (e.g., Jeff Geerling's GitHub) and package sources (PyPI) are well-known technology services.

ansible-validator