google-styleguides-skills
Pass
Audited by Gen Agent Trust Hub on Feb 25, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill package is composed entirely of Markdown files containing coding conventions and metadata. There is no executable logic, scripts, or binary files present.
- [PROMPT_INJECTION]: The content defines coding 'Golden Rules' and standards (e.g., naming, formatting, and documentation). These instructions are specific to code style and do not attempt to bypass agent safety filters or override core behavioral directives.
- [DATA_EXFILTRATION]: No network operations, data collection logic, or sensitive file access patterns were identified. The skill is entirely self-contained documentation.
- [REMOTE_CODE_EXECUTION]: There are no commands or configurations that download and execute external scripts. The installation instructions provided in the README are standard for the Kiros/Skills ecosystem and point to the official vendor repository.
Audit Metadata