kql
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acts as a translator for user-supplied data (such as SQL or Splunk queries), which introduces a potential surface for indirect prompt injection. The skill includes specific instructions for the agent to follow structured .kql file formats and use service-specific context, which helps maintain operational boundaries when processing untrusted input.
- Ingestion points: User queries and descriptions provided in the primary prompt to be translated or optimized.
- Boundary markers: Explicit instructions in SKILL.md and workflows/WriteQuery.md to use a defined header format and inline comments for generated output.
- Capability inventory: The skill provides instructions for generating KQL code, CLI commands, and Python scripts but does not contain local scripts to execute them automatically.
- Sanitization: Relies on the host environment's default safety guardrails and the agent's internal routing logic.
- [COMMAND_EXECUTION]: The workflows/Tooling.md file provides practical examples of shell commands for Azure CLI, PowerShell, and jq to help users run and export queries. These commands are standard administrative tools and are presented for instructional purposes rather than for direct, autonomous execution by the skill itself.
Audit Metadata