The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses Python scripts to perform local audits and initialization of configuration files. These commands are restricted to the local environment and intended for the skill's primary purpose.
[DYNAMIC_EXECUTION]: The script scripts/git4d_schema_audit.py includes a safe check for the yaml library, using yaml.safe_load() if available, which prevents arbitrary code execution during YAML parsing. It also provides a conservative fallback parser for environments without PyYAML.
[DATA_EXPOSURE]: The skill requires full repository read access to audit configurations, which is appropriate for its stated purpose. Write access is restricted to specific directories (./git4d-results/, ./artifacts/, ./reports/).
[EXTERNAL_DOWNLOADS]: The skill references official documentation and repositories from trusted domains (e.g., GitHub, json-schema.org, yaml.org). These are documented neutrally and do not involve remote code execution.

git4d-schema