The Agent Skills Directory

[DATA_EXFILTRATION]: The skill identifies and reads the dbt profiles.yml configuration file to resolve environment context.
Evidence: scripts/sandbox/parse_profiles.py reads the file to extract database, schema, and role information used for environment classification.
[COMMAND_EXECUTION]: The skill executes local dbt commands to build models in a development sandbox.
Evidence: Workflow 4.1 in SKILL.md describes running dbt build --select <model> after performing environment classification and receiving user confirmation.
[COMMAND_EXECUTION]: The skill includes Python test scripts that execute helper utilities via the shell for verification.
Evidence: Multiple test files in scripts/sandbox/tests/ use subprocess.run() to call utility scripts such as classify_sandbox.py and parse_profiles.py during functional testing.
[EXTERNAL_DOWNLOADS]: The documentation references the installation of vendor-provided packages and CLI tools.
Evidence: README.md instructs the user to install the montecarlodata package via pip and the skill via npx.

monte-carlo-prevent