exasol-transformers

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides documentation and Python/SQL examples for the Exasol Transformers Extension. It utilizes the vendor's own libraries (exasol.nb_connector.transformers_extension_wrapper) to perform infrastructure tasks like deploying containers to BucketFS and installing UDF scripts.
  • [CREDENTIALS_UNSAFE]: The skill correctly instructs the user to store sensitive values (DB passwords, BucketFS passwords, Hugging Face tokens) in a secure configuration store (my_secrets) and does not hardcode any credentials or output them to the user.
  • [EXTERNAL_DOWNLOADS]: The skill mentions downloading and installing models from Hugging Face using standard tools (transformers, install_model). These operations target a well-known service (Hugging Face) for their intended purpose in an NLP extension.
  • [COMMAND_EXECUTION]: The skill generates SQL commands (get_activation_sql, UDF calls) and Python calls for database interaction. These are standard operations for a database extension and do not involve arbitrary shell command execution or privilege escalation.
  • [DATA_EXFILTRATION]: No data exfiltration patterns were detected. Network operations (Hugging Face model downloads) are part of the stated functionality and target trusted domains.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 04:31 PM
Security Audit — agent-trust-hub — exasol-transformers