accounting-finance-advisory
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of instructional text and documentation templates. It does not contain any scripts, command-line executions, or automated network operations. Author attribution links are benign.
- [INDIRECT_PROMPT_INJECTION]: The skill manages the ingestion of external data such as client ledgers and project ToRs, which presents a surface for indirect prompt injection. The risk is minimized by explicit boundary constraints and validation steps. 1. Ingestion points: Client finance records and ToR documents (SKILL.md). 2. Boundary markers: Workflow interruption rules for missing data and a read-only review policy (SKILL.md). 3. Capability inventory: Read and search access only (SKILL.md). 4. Sanitization: Mandatory verification against canonical doctrine and authoritative sources (SKILL.md).
Audit Metadata