The Agent Skills Directory

[SAFE]: The skill serves as a documentation guide for the llamafarm-common package. It encourages secure authentication practices by recommending the use of environment variables (HF_TOKEN) rather than hardcoding secrets.
[SAFE]: The instructions explicitly warn against high-severity security anti-patterns, such as logging authentication tokens, and provide secure alternatives for handling sensitive data.
[SAFE]: Guidelines for model identification include sanitization steps to strip potentially problematic suffixes from model IDs before they are used in API calls to the HuggingFace Hub.
[SAFE]: The skill promotes safe testing methodologies, requiring the use of mocks for all external API calls and temporary directories for file system operations to ensure network isolation and environment cleanliness.
[SAFE]: Network operations described in the patterns target well-known and trusted services (HuggingFace Hub) using official SDKs, with no evidence of obfuscation, persistence mechanisms, or unauthorized privilege escalation.

common-skills