monitoring-observability

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill contains well-structured and industry-standard configurations for observability without malicious patterns.
  • [DATA_EXFILTRATION]: The skill defines constraints that explicitly forbid logging sensitive data like passwords and API keys.
  • [EXTERNAL_DOWNLOADS]: Links refer to official documentation for Prometheus and Grafana, which are trusted and safe resources.
  • [PROMPT_INJECTION]: The logging middleware ingests untrusted User-Agent strings and request paths. Ingestion points: User-Agent and path in Step 3 logging middleware in SKILL.md. Boundary markers: None. Capability inventory: File-writing via Winston transports in SKILL.md. Sanitization: Uses JSON formatting to prevent basic log forging.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 01:50 PM