cx-data-pipeline
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill facilitates standard administrative tasks for the Coralogix platform using the vendor-provided cx CLI tool. All workflows, including parsing rule creation and Events2Metrics configuration, are consistent with legitimate platform management.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from log queries and local JSON files. Analysis shows this surface is inherent to the skill's purpose and does not incorporate patterns designed to exploit the agent.
- Ingestion points: Data enters the context via
cx logsand configuration files loaded through--from-fileinSKILL.md. - Boundary markers: The instructions recommend using existing resources as templates to maintain structure.
- Capability inventory: Tools are limited to the Coralogix CLI and
jqfor JSON manipulation. - Sanitization: Input validation is handled by the underlying CLI and API.
- [PROMPT_INJECTION]: Static detections of concealment are confirmed to be false positives. Phrases interpreted as hiding actions are technical guidance on platform behavior, such as warnings about ingestion delays in archive storage and potential metric name collisions during aggregation.
Audit Metadata