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.mdis 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