The Agent Skills Directory

[SAFE]: The skill instructions and associated reference documents do not contain any malicious code, obfuscation, or unauthorized data exfiltration patterns. All operations are aligned with the stated purpose of managing semantic layer definitions.
[COMMAND_EXECUTION]: The skill utilize standard CLI tools such as revos, bq, and dbt for metadata discovery and validation. These commands are restricted to the local environment and the user's database dataset, which is consistent with project management workflows.
[SAFE]: The skill ingests data from local dbt models and BigQuery schemas. No untrusted remote data sources are accessed, and no external code is downloaded or executed.

create-cubes