hugging-face-cli
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
hfCLI to perform repository management tasks, such as creating, deleting, and tagging repositories on the Hugging Face Hub. These operations are standard for the tool and execute within the user's authenticated session. - [EXTERNAL_DOWNLOADS]: Commands like
hf downloadfacilitate fetching models and datasets from Hugging Face's official servers. These interactions target a well-known technology service and align with the skill's primary purpose. - [REMOTE_CODE_EXECUTION]: The skill includes instructions for
hf jobs run, which allows users to execute compute jobs on Hugging Face's managed cloud infrastructure. This is a primary feature of the service and is used for legitimate machine learning workflows. - [CREDENTIALS_UNSAFE]: Authentication practices described in the skill follow security best practices by utilizing environment variables (e.g.,
$HF_TOKEN) and the CLI's internal secret management rather than hardcoding sensitive credentials.
Audit Metadata