ai-ai-observability
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as a comprehensive architectural and implementation guide for AI observability. It covers various levels of scale and budget, providing appropriate tool recommendations for each.
- [COMMAND_EXECUTION]: The skill provides code examples for instrumenting applications with observability SDKs (Python and pseudo-code). These examples follow best practices, such as using environment variables for sensitive configurations and implementing PII redaction before data export.
- [EXTERNAL_DOWNLOADS]: The skill references well-known and trusted observability platforms such as LangSmith, LangFuse, Datadog, and Helicone. All referenced repositories (e.g., anthropics/skills) are from trusted organizations or well-known services.
- [DATA_EXFILTRATION]: There is no evidence of malicious data exfiltration. The instructions explicitly warn against storing PII in observability backends and provide technical methods for hashing and redaction at the collector level.
- [CREDENTIALS_UNSAFE]: The skill adheres to secure credential management practices, explicitly instructing users to never hardcode API keys and to use environment variables instead.
Audit Metadata