phoenix-tracing
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill content is purely educational and technical documentation. It provides comprehensive reference material for developers to implement observability in LLM applications.
- [EXTERNAL_DOWNLOADS]: The documentation references standard and official libraries from the vendor (arize-ai) and the OpenTelemetry project for instrumentation. These are reputable sources intended for monitoring application performance and behavior.
- [DATA_EXFILTRATION]: The guide describes how to transmit trace data to a Phoenix observability server. This is a core feature of the platform. Importantly, the documentation includes specific instructions for production security, such as using environment variables to mask PII (Personally Identifiable Information) and sensitive data before it is exported.
- [INDIRECT_PROMPT_INJECTION]: The skill documents the ingestion of external data (agent trajectories, LLM traces). This creates a potential surface for indirect prompt injection at the platform level, but the skill itself focuses on implementation and includes specific guidance on data masking and sanitization.
Audit Metadata