copilot-studio-harness-picker

Pass

Audited by Gen Agent Trust Hub on Aug 30, 2026

Risk Level: SAFE
Full Analysis
  • Secure Local Computation: The skill includes a Python script for credit estimation that operates locally using only the standard library. It performs no network activity and includes validation logic to ensure data integrity during arithmetic operations.
  • Sensitive Data Protection: Skill instructions provide explicit guardrails that prevent the agent from requesting passwords, connection strings, or production secrets, instructing it to rely on sanitized descriptions instead.
  • Vendor Trust Alignment: All external resources and documentation links point to official Microsoft domains or the official Microsoft GitHub organization, aligning with the trusted status of the vendor.
  • Clear Architectural Boundaries: The decision-making logic separates confirmed requirements from assumptions, ensuring that recommendations are based on verifiable evidence and dated official sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 30, 2026, 11:55 PM
Security Audit — agent-trust-hub — copilot-studio-harness-picker