break-trace

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFE
Full Analysis
  • Internal Data Retrieval: The skill is designed to fetch journal entries and transaction details from internal financial systems. The instructions in SKILL.md guide the agent to use the internal-gl MCP and subledger MCP to pull data for comparison.
  • Diagnostic Safeguards: The skill includes explicit instructions that limit its role to analysis rather than execution of changes. The documentation in SKILL.md clarifies that the skill only diagnoses and does not post adjustments, ensuring a separation between analysis and data modification.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 05:26 PM
Security Audit — agent-trust-hub — break-trace