in-product-messaging
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing external data from sources such as survey answers, tickets, and web pages (SKILL.md). This risk is mitigated by explicit safety instructions within the skill that act as boundary markers (treating input as data, not instructions) and sanitization protocols (requiring human confirmation before executing sensitive tasks like bulk updates or sending messages).
- [EXTERNAL_DOWNLOADS]: The skill references documentation and legal guidelines from trusted organizations including Apple, Google, the UK Information Commissioner's Office (ICO), and official government legislative sites for research purposes.
Audit Metadata