python-observability
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains technical documentation and code examples for observability using well-known, legitimate Python libraries such as structlog, prometheus_client, and OpenTelemetry.
- [DATA_EXFILTRATION]: While the skill demonstrates network operations (using httpx and OTLP exporters), these are limited to the intended purpose of propagating correlation IDs and exporting telemetry data to user-defined endpoints. No sensitive file access or hardcoded credentials were found.
- [REMOTE_CODE_EXECUTION]: No patterns for remote code execution or untrusted package installation were detected. All code snippets use standard APIs for logging and metrics collection.
- [PROMPT_INJECTION]: The skill contains no instructions aimed at overriding agent behavior or bypassing safety filters.
- [INDIRECT_PROMPT_INJECTION]: The skill provides patterns for logging request data (IDs, paths, errors). While this involves ingesting untrusted data, the skill is focused on machine-readable logging (JSON) and does not involve the agent acting upon or executing instructions found within that data.
Audit Metadata