export-control-intake

Pass

Audited by Gen Agent Trust Hub on Jul 9, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources to generate reports, creating a surface for potential indirect prompt injection attacks.
  • Ingestion points: SKILL.md identifies inputs from "Read" operations on external files such as "研究計画" (research plans), "契約案" (contract drafts), and "装置仕様" (equipment specifications).
  • Boundary markers: The instructions do not define explicit delimiters or instructions to ignore embedded commands within the processed documents.
  • Capability inventory: The skill references capabilities for Write, WebSearch, WebFetch, and Bash, which provide significant actions that could be targeted via injection.
  • Sanitization: No validation or sanitization requirements are specified for the content of the ingested materials.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 9, 2026, 03:37 AM
Security Audit — agent-trust-hub — export-control-intake