The Agent Skills Directory

Secure YAML Parsing Implementation: The scripts/validate_catalog.py script uses a specialized YAML loader derived from yaml.SafeLoader. It is designed to proactively reject potentially problematic YAML features, such as aliases, anchors, and custom tags, which helps ensure that processed data remains within safe parameters. While a deterministic scan noted the use of the load function, the implementation context demonstrates a secure approach.
File System Protection: The skill includes input validation that prevents path traversal by checking for parent directory segments ('..') and ensuring files are within permitted directories. It also enforces a maximum input size limit, which provides a layer of protection against resource-intensive processing.
Data Integrity and Schema Enforcement: By utilizing JSON Schema for structural validation, the skill ensures that all catalog entries strictly adhere to a predefined contract. This approach minimizes the risk of processing unexpected or malformed data structures during the catalog enrichment workflow.

data-catalog