seo-keyword-research
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by crawling external websites to extract SEO context.
- Ingestion points:
scripts/site_crawler.pyfetches content from user-provided URLs. - Boundary markers: The skill uses specific system prompts to guide the LLM's interpretation of crawled data, though explicit data delimiters are not used in the sample scripts.
- Capability inventory: The skill is limited to file output (for reports) and API calls to the AIsa service. It does not possess capabilities for arbitrary command execution or unauthorized file system access.
- Sanitization: Content is normalized and cleaned of HTML tags and excessive whitespace before being processed by the AI.
Audit Metadata