data-quality-frameworks
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill includes instructions to install the
great_expectationslibrary using a package manager. This is a well-known and trusted library for data quality validation. - [COMMAND_EXECUTION]: The playbook provides common command-line instructions for initializing and configuring data quality projects, such as
great_expectations initandgreat_expectations datasource new. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and validate data from external warehouses (e.g., Snowflake, as seen in
orders_contract.yaml). This ingestion of untrusted external content represents a surface for indirect prompt injection. - Ingestion points: External data assets and warehouse tables processed by Great Expectations and dbt (referenced in
quality_pipeline.pyandorders_contract.yaml). - Boundary markers: The provided code templates do not define explicit boundary markers or delimiters for processed data.
- Capability inventory: The skill utilizes CLI tool execution and network-based alerting (e.g., Slack webhooks mentioned in
orders_checkpoint.yml). - Sanitization: While the skill advises handling sensitive data securely, no specific sanitization logic for prompt injection is included in the implementation examples.
Audit Metadata