The Agent Skills Directory

[SAFE]: The skill provides architectural advice and code snippets for Machine Learning operations. It follows best practices for feature store design, such as point-in-time joins and feature validation.
[EXTERNAL_DOWNLOADS]: The documentation references standard Python libraries such as feast, great_expectations, pandera, numpy, and pandas. These are well-known and trusted packages within the data science and ML community.
[DATA_EXFILTRATION]: The skill describes connecting to common data sources and sinks (S3, GCS, BigQuery, Kafka, Redis) for feature processing. These operations are essential to the skill's stated purpose and do not involve unauthorized data movement.
[INDIRECT_PROMPT_INJECTION]: While the skill involves processing external data from sources like Kafka and S3, it incorporates robust validation logic using great_expectations and pandera to ensure data quality and schema adherence, which serves as a security best practice when handling external inputs.

ml-feature-store