The Agent Skills Directory

[SAFE]: The skill provides legitimate technical documentation for configuring and using DataHub for metadata management.\n- [EXTERNAL_DOWNLOADS]: The skill instructions include installing the official DataHub Python package (acryl-datahub) and adding the official DataHub Helm repository (https://helm.datahubproject.io/). These are well-known and expected sources for the tool.\n- [SAFE]: The skill demonstrates security best practices by recommending environment variable substitution (e.g., ${DATAHUB_TOKEN}, ${AWS_SECRET_ACCESS_KEY}) for managing secrets within ingestion recipes.\n- [PROMPT_INJECTION]: The skill facilitates the ingestion of metadata from various external sources (databases, S3, Kafka), which represents a surface for indirect prompt injection.\n- Ingestion points: Metadata is ingested via YAML recipes from PostgreSQL, Hive, Kafka topics, S3 files, and dbt artifacts.\n- Boundary markers: The instructions do not define specific delimiters for separating ingested metadata from system instructions.\n- Capability inventory: The skill utilizes the datahub CLI, Python SDK emitters, and REST/GraphQL APIs for processing and storing metadata.\n- Sanitization: Content sanitization is not explicitly discussed, as the focus is on the technical ingestion mechanism. This surface is an inherent characteristic of the tool's intended use case.

datahub-catalog