taxonomy

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill establishes a robust security posture by requiring explicit user confirmation for all write operations and prohibiting autonomous destructive actions.
  • [SAFE]: All tool operations are scoped to the Amplitude platform's internal libraries and official MCP tools, ensuring no unauthorized external data transmission.
  • [SAFE]: The instructions prioritize data integrity and safety, specifically warning against sampling or unilateral deletion which could impact downstream analytics.
  • [SAFE]: No evidence of prompt injection, obfuscated content, or malicious code execution patterns was found in the skill documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 07:04 AM
Security Audit — agent-trust-hub — taxonomy