scrapebox-seo-automation-toolkit
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPRIVILEGE_ESCALATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines workflows for harvesting keywords and scraping content from external search engines and competitor websites. This untrusted data could contain malicious instructions meant to manipulate an AI agent during analysis.
- Ingestion points: Data enters the workflow through the 'Keyword Scraper', 'Backlink Checker', and 'Content Scraping' components (SKILL.md).
- Boundary markers: The instructions and provided Python processing scripts do not define clear delimiters or warnings to ignore embedded instructions in the scraped content.
- Capability inventory: The skill provides Python scripts that parse, filter, and alert on data exported from these scrapers.
- Sanitization: There is no evidence of sanitization, validation, or escaping of the external content before it is processed by the logic.
- [PRIVILEGE_ESCALATION]: The 'Setup Process' section instructs the user to run the application installer with administrator privileges on Windows. While common for desktop software installation, this requires the acquisition of elevated permissions.
Audit Metadata