phoenix-observability
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing the
arize-phoenixecosystem and OpenTelemetry instrumentors from public package registries. These are established tools in the AI observability field. - [COMMAND_EXECUTION]: Documentation includes standard shell commands for server management, including service startup, port configuration, and container deployment. These actions are consistent with the platform's administrative requirements.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a natural attack surface as it is designed to ingest and evaluate LLM inputs and outputs.
- Ingestion points: Data is collected through OpenTelemetry instrumentation and queried via the Phoenix Client API (SKILL.md).
- Boundary markers: External content is interpolated into evaluation templates for assessment (references/advanced-usage.md).
- Capability inventory: The skill performs network operations to logging endpoints, database writes for persistence, and local file exports for backups.
- Sanitization: Model outputs are processed by evaluators without explicit sanitization, which is standard behavior for observability and quality assessment tools.
Audit Metadata