seo-programmatic

Pass

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze data from various external and local sources, creating a potential surface for indirect prompt injection if those sources contain malicious instructions.
  • Ingestion points: Processes data from CSV/JSON files, API endpoints, database queries, and local files within the .seo-cache/ directory (SKILL.md).
  • Boundary markers: Includes instructions to parse only clearly valid fields from the cache and implement quality gates based on word count and uniqueness (SKILL.md).
  • Capability inventory: Performs file read/write operations to the shared data cache and generates content templates based on the analyzed data (SKILL.md).
  • Sanitization: Employs data quality checks for record uniqueness and field completeness, as well as a 40% uniqueness threshold to differentiate generated content, which helps mitigate automated injection patterns.
  • [SAFE]: The skill explicitly instructs the user/agent to add the .seo-cache/ directory to .gitignore. This is a security best practice that prevents sensitive analysis data or cached metadata from being accidentally committed to version control systems.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 19, 2026, 07:39 AM
Security Audit — agent-trust-hub — seo-programmatic