awsflow-glue
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection due to its interaction with external metadata.
- Ingestion points: The skill retrieves metadata from the AWS Glue Data Catalog via commands like GetTable, GetDatabase, and GetPartitions in SKILL.md.
- Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish between its primary instructions and metadata retrieved from the catalog.
- Capability inventory: The skill has the ability to create, update, and delete AWS resources using commands such as CreateJob, UpdateJob, DeleteJob, and DeleteTable.
- Sanitization: The skill does not define any mechanisms for sanitizing or validating the metadata retrieved from AWS Glue APIs before it is incorporated into the agent's context.
Audit Metadata