seo-drift

Pass

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches HTML content from user-provided URLs to analyze SEO elements. This is done through a centralized script (fetch_page.py) that incorporates SSRF protection by blocking private, loopback, and reserved IP ranges via google_auth.validate_url().
  • [COMMAND_EXECUTION]: Orchestrates SEO monitoring tasks by executing a suite of local Python scripts (e.g., drift_baseline.py, drift_compare.py, drift_report.py). It also generates an HTML report locally on the filesystem.
  • [PROMPT_INJECTION]: Ingests and processes untrusted HTML data from external websites, creating a surface for indirect prompt injection. The skill manages this by treating the data as structured input for a set of 17 predefined comparison rules and storing elements in a SQLite database rather than directly interpolating raw HTML into agent prompts.
  • Ingestion points: External URLs processed by scripts/fetch_page.py and scripts/parse_html.py.
  • Boundary markers: Not explicitly defined for the prompt structure.
  • Capability inventory: SQLite database write access, local file system write (HTML reports), and network fetch capabilities.
  • Sanitization: Implements SSRF validation for all target URLs.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 15, 2026, 05:45 PM