programmatic-seo

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes data from external SEO tools and search engine results, creating an attack surface where third-party content can influence the agent.
  • Ingestion points: Keyword patterns, demand trends, and live SERP evidence retrieved via the DataForSEO API tools listed in references/sandbase-api-map.md.
  • Boundary markers: The skill requires user authorization for file writes and publishing but lacks delimiters to isolate external tool outputs from system instructions.
  • Capability inventory: The skill can generate and write files and perform network operations via the sandbase_call_tool function.
  • Sanitization: The instructions do not define methods to filter or validate external API content before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 08:11 AM
Security Audit — agent-trust-hub — programmatic-seo