skills/databricks-solutions/lakebase-online-ml/instrumenting-with-mlflow-tracing/Gen Agent Trust Hub
instrumenting-with-mlflow-tracing
Pass
Audited by Gen Agent Trust Hub on Aug 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides legitimate documentation and verification steps for MLflow instrumentation, authored by databricks-solutions.
- [PROMPT_INJECTION]: No evidence of prompt injection patterns, role-play bypasses, or instructions to ignore safety filters. The primary purpose is strictly technical instrumentation.
- [DATA_EXFILTRATION]: No sensitive file access or unauthorized network exfiltration detected. The skill explicitly mentions PII redaction as a security practice for advanced tracing patterns.
- [CREDENTIALS_UNSAFE]: No hardcoded API keys or secrets were found. The documentation correctly points to environment variables for production configuration.
- [COMMAND_EXECUTION]: The provided code snippets are limited to standard MLflow library calls for trace verification and do not involve unsafe shell execution or subprocess spawning.
Audit Metadata