blog-locale-audit
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from blog posts, frontmatter fields, and JSON configuration files which are treated as untrusted data. The instructions proactively mitigate the risk of indirect prompt injection by explicitly commanding the agent to extract values only and ignore any instructions found within the audited files.
- [SAFE]: The skill includes path resolution and validation logic (e.g., rejecting symlinks and traversal outside the root) to prevent directory traversal attacks when identifying target directories.
- [SAFE]: For HTML report generation, the skill mandates the use of
html.escapeon all dynamic values, which prevents potential Cross-Site Scripting (XSS) vulnerabilities in the output. - [SAFE]: External references are restricted to well-known documentation from Google (developers.google.com) regarding internationalization standards.
Audit Metadata