aws-architecture
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Prompt Injection] (SAFE): No instructions attempting to override agent behavior or bypass safety filters were detected. The content consists of architectural advice.
- [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file paths, or network exfiltration patterns were found. The skill does not perform any network operations.
- [Obfuscation] (SAFE): The content is standard markdown with no hidden characters, Base64 encoding, or homoglyphs.
- [Unverifiable Dependencies & Remote Code Execution] (SAFE): The skill does not install any packages or download/execute remote scripts. The CDK and CloudFormation examples are static code snippets for documentation purposes only.
- [Privilege Escalation] (SAFE): No commands related to privilege escalation (e.g., sudo, chmod) are present.
- [Persistence Mechanisms] (SAFE): No logic exists to modify system configurations or create persistent tasks.
- [Dynamic Execution] (SAFE): The skill contains no runtime execution logic, such as eval() or subprocess calls, that would process untrusted input.
Audit Metadata