regulated-domain-triggers
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from the project environment, including source code, schemas, routes, and UI copy, creating a surface for indirect prompt injection. However, the risk is mitigated by explicit instructions that prioritize human oversight and forbid the agent from making definitive safety claims. Ingestion points: The skill searches through product copy, schemas, dependencies, and metadata as described in SKILL.md. Boundary markers: The instructions do not define specific delimiters for separating ingested data from the prompt. Capability inventory: The skill is restricted to reading information and writing to a specific context file (.readyvibe/context.md), lacking capabilities for arbitrary shell execution or network exfiltration. Sanitization: No explicit sanitization or filtering of external data is specified.
- [EXTERNAL_DOWNLOADS]: The skill references a comprehensive list of official regulatory and standards-body domains in references/official-sources.md. These references target well-known and trusted organizations, such as the Federal Trade Commission (ftc.gov), the European Data Protection Board (edpb.europa.eu), and the World Wide Web Consortium (w3.org), for the purpose of legitimate information lookup and compliance research.
Audit Metadata