context-engineering
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of instructional text guiding users on how to structure information for AI agents. No executable code or malicious commands are present.
- [SAFE]: The content explicitly recommends security best practices, such as preventing the commitment of secrets, .env files, or private keys to repositories.
- [SAFE]: It addresses the risk of indirect prompt injection by identifying 'Untrusted' context sources (third-party API responses, external docs) and instructing agents to treat instruction-like content in those sources as data rather than directives.
Audit Metadata