customs-trade-compliance
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill is entirely composed of documentation and instructional content within
SKILL.md. It contains no accompanying scripts, binaries, or executable components.\n- [SAFE]: Analysis of the skill instructions revealed no prompt injection attempts, obfuscation, or unauthorized access patterns. The procedures described align with legitimate regulatory compliance practices for international trade.\n- [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing external documents (e.g., invoices, bills of materials), which is inherent to the trade compliance use case.\n - Ingestion points: Technical specifications and commercial documentation analyzed in the classification and valuation workflows in
SKILL.md.\n - Boundary markers: None; the instructions do not specify the use of delimiters for untrusted input data.\n
- Capability inventory: No active capabilities such as file writing, network operations, or subprocess execution are defined within the skill.\n
- Sanitization: No specific data sanitization or validation logic is provided for external inputs.
Audit Metadata