views-and-navigation
Pass
Audited by Gen Agent Trust Hub on May 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of educational documentation and Java code snippets for the Vaadin web framework. No executable scripts, system modifications, or unauthorized network operations were detected.
- [PROMPT_INJECTION]: The instructions do not contain patterns intended to bypass safety guidelines, override agent behavior, or extract system prompts.
- [DATA_EXFILTRATION]: No access to sensitive local files (such as SSH keys, AWS credentials, or environment files) is requested or performed. The skill does not communicate with external domains.
- [EXTERNAL_DOWNLOADS]: The skill does not include any external dependencies, package installations, or remote script executions.
- [OBFUSCATION]: No hidden text, Base64-encoded commands, or homoglyph-based obfuscation techniques were found in the provided files.
- [INDIRECT_PROMPT_INJECTION]: While the skill teaches how to handle URL and query parameters (standard web development practice), it does not create a vulnerability surface for the agent itself. The skill focuses on building UI components and navigation logic.
Audit Metadata