approve-exempt
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No security issues detected. The skill follows best practices for AI agent interactions with sensitive platform operations.
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it retrieves and displays data (exemption titles, reasons) that could be populated by users. However, this is managed by a 'list-then-select' workflow and explicit user confirmation steps, effectively mitigating the risk of the agent autonomously following instructions embedded in the data.
- Ingestion points: harness_list tool in SKILL.md
- Capability inventory: harness_execute tool in SKILL.md
- Boundary markers: Data is formatted into a markdown table.
- Sanitization: Relies on platform-level validation and human-in-the-loop confirmation.
Audit Metadata