user-onboarding-sop
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill defines a routine documentation task for creating an AGENTS.md file to assist end-users with setup and usage.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies a surface for indirect injection via the web search discovery phase. 1. Ingestion points: Web search results for official documentation (SKILL.md). 2. Boundary markers: None defined. 3. Capability inventory: Limited to markdown text generation for documentation. 4. Sanitization: The instructions mandate using only verified URLs and expressly forbid inventing content (SKILL.md).
Audit Metadata