write-aql
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted natural language requests to generate AQL queries and ingests external data such as dataset schemas and data previews.
- Ingestion points: User analytics requests, dataset fields/models, and sample data from preview queries (SKILL.md).
- Boundary markers: Absent.
- Capability inventory: Generation, validation, and execution of queries via the
anfratool and associated skills (run-aql). - Sanitization: Absent.
- [DYNAMIC_EXECUTION]: The skill dynamically generates AQL (Analytics Query Language) code from user input and metadata. This generated code is executed via the
run-aqltool to retrieve results. This behavior is the primary intended function of the skill and uses standard query generation patterns.
Audit Metadata