query-app-insights
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
tsx scripts/queryAppInsights.ts "<kql-query>"to execute queries against the Azure Application Insights API. While these are controlled shell commands within the agent's environment, they allow for arbitrary KQL (Kusto Query Language) execution. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process telemetry data from an external source (Azure Application Insights). This creates a surface where an attacker who can influence the telemetry data (e.g., by triggering specific events with malicious payloads in custom dimensions) could attempt to influence the agent's interpretation of that data.
- Ingestion points:
scripts/queryAppInsights.tsfetches data from the Application Insights API, which is then presented to the agent. - Boundary markers: None explicitly mentioned in the instructions for the agent when interpreting the JSON results.
- Capability inventory: The agent can execute shell commands via
tsxto query the API. - Sanitization: The script performs standard JSON parsing but does not sanitize the content of the telemetry values for potential prompt injection before the agent processes them.
Audit Metadata