audit-support-master

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely instructional, serving as a template for the AI agent to follow during financial audit coordination. It does not contain any malicious patterns or hidden instructions.
  • [NO_CODE]: No scripts, external packages, or shell commands are included. The functionality relies on prompting and calls to other defined agent skills.
  • [DATA_EXPOSURE]: While the skill handles sensitive financial audit metadata (risk levels, adjustment amounts), it does not contain any tools or instructions for network exfiltration or unauthorized file access. Access to ../../CLAUDE.md is a localized configuration read standard for this type of agent framework.
  • [PROMPT_INJECTION]: No evidence of prompt injection or bypass attempts was found. The instructions are clearly scoped to audit support tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 02:25 AM
Security Audit — agent-trust-hub — audit-support-master