box-legal-workflows-intake

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references prerequisites to be installed from an official repository (github.com/box/box-for-ai).
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze third-party intake documents, which represents an attack surface for indirect prompt injection.
  • Ingestion points: External documents stored in specific intake folders analyzed by tools like ai_qa_multi_file.
  • Boundary markers: None explicitly defined in this specific instruction file; the skill relies on underlying MCP tool implementations.
  • Capability inventory: Folder inventory, metadata management, file commenting, collaboration sharing, and DocGen document creation.
  • Sanitization: Includes strict instructions to surface facts and citations only, prohibiting the agent from performing conflict or risk determinations.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 06:16 AM
Security Audit — agent-trust-hub — box-legal-workflows-intake