anfra-run-aql
Pass
Audited by Gen Agent Trust Hub on Nov 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: Shell commands are constructed using dynamic placeholders for dataset names and AQL queries. * Evidence: anfra query --dataset '' in SKILL.md. * Description: The use of single quotes for the query parameter in the command examples suggests an attempt at containment, but the agent may still be vulnerable to shell injection if user-provided strings are not strictly validated or escaped before execution.
- [INDIRECT_PROMPT_INJECTION]: The skill retrieves data from external datasets that may contain untrusted content, creating a surface for indirect instructions. * Ingestion points: Data rows fetched via the anfra query command (SKILL.md). * Boundary markers: No explicit boundary markers or instructions to treat returned data as non-executable text are provided in the skill content. * Capability inventory: The agent uses query results to perform data analysis, including reporting counts, sums, or summaries. * Sanitization: No sanitization or validation of dataset content is performed before the agent processes the results.
Audit Metadata