newrelic-cli-diagnostics
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using the
newrelicCLI andjqutility to perform diagnostic runs, validate configurations, and query telemetry data via NRQL. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources, such as agent configuration files and API responses, which constitutes an indirect injection attack surface.
- Ingestion points: Reads local agent configuration files (
newrelic.yml,newrelic.config) and fetches application metadata and transaction counts from New Relic's APIs. - Boundary markers: No explicit boundary markers or instructions to ignore embedded instructions are present in the processing steps.
- Capability inventory: Includes the ability to read configuration files and perform network operations via the CLI.
- Sanitization: The skill relies on the default behavior of the
newrelicCLI andjq; no explicit sanitization of ingested content is defined in the instructions.
Audit Metadata