meta-tags-optimizer
Pass
Audited by Gen Agent Trust Hub on May 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious code, obfuscation, or hardcoded credentials were found in any of the skill's components.
- [EXTERNAL_DOWNLOADS]: The skill links to supplementary documentation, formulas, and HTML templates hosted on the author's official GitHub repository (
github.com/aaron-he-zhu/seo-geo-claude-skills). These are vendor-owned resources used for instructional support. - [PROMPT_INJECTION]: The skill is designed to process external web content or URLs, which creates a potential surface for indirect prompt injection.
- Ingestion points: The skill accepts a
<page URL or content>argument as specified inSKILL.md. - Boundary markers: Instructions do not explicitly define delimiters for untrusted content.
- Capability inventory: The skill is authorized to write metadata deliverables to the
memory/directory. - Sanitization: No specific sanitization or filtering logic is described in the workflow.
- Context: While the surface exists, the risk is managed by the skill's specific focus on SEO metadata generation and its lack of high-privilege capabilities.
- [COMMAND_EXECUTION]: The workflow involves saving results to local markdown files (e.g.,
memory/content/,memory/hot-cache.md) to maintain state. These operations are consistent with standard agent memory management and do not involve arbitrary shell command execution.
Audit Metadata