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
Risk Level
SAFE
Analyzed
Jun 13, 2026, 09:10 AM
Security Audit — agent-trust-hub — ai-ai-observability