programmatic-seo

Pass

Audited by Gen Agent Trust Hub on Feb 21, 2026

Risk Level: SAFE
Full Analysis
  • [Prompt Injection] (SAFE): The instructions follow a standard expert persona for SEO strategy. No attempts to bypass safety filters or override system instructions were found.
  • [Data Exposure & Exfiltration] (SAFE): No hardcoded secrets, sensitive system paths, or network exfiltration patterns were detected. It references a local marketing context file which is standard for the environment.
  • [Unverifiable Dependencies & Remote Code Execution] (SAFE): No package installations or remote script executions are present. The skill consists of markdown documentation.
  • [Indirect Prompt Injection] (LOW): The skill mentions using external data sources like scraped data for populating SEO templates. This represents a potential ingestion surface for untrusted data, though the skill itself does not provide the execution logic to process it.
  • Ingestion points: External/scraped data sources (referenced in Implementation Framework).
  • Boundary markers: None specified for data interpolation.
  • Capability inventory: None; the skill is documentation-only.
  • Sanitization: None specified for external content.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 21, 2026, 10:26 AM