seo-ai-visibility
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection Surface: The skill is designed to fetch and process content from external, untrusted websites, which is a potential vector for indirect prompt injection.
- Ingestion points: The skill uses the WebFetch tool to read content from public websites, including homepages, service pages, and metadata files like robots.txt (SKILL.md, Step 1).
- Boundary markers: The skill explicitly references a shared safety protocol (untrusted-content.md) and includes a section that instructs the agent to treat all external text as data rather than commands (SKILL.md).
- Capability inventory: The skill can read external sites, generate HTML artifacts, and write to connected platforms like Wix or Shopify (SKILL.md, Step 5 and 6).
- Sanitization: The skill requires user verification for all facts and implements a batch approval process for any automated writes to external platforms (reference/catalog_refresh.md).
- External Platform Integration: The skill can connect to and modify content on platforms like Shopify and Wix.
- Context: These capabilities are used to apply SEO fixes directly. The skill mitigates risk by requiring explicit user approval for every batch of changes and providing a manual CSV-based alternative (SKILL.md, Step 6).
Audit Metadata