content-management-systems

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill contains only documentation and workflow instructions for CMS development, with no executable code or scripts provided.
  • [SAFE]: The instructions explicitly recommend security best practices, such as maintaining CSRF safeguards and implementing validation rules when changing editor or admin behavior.
  • [NO_CODE]: The skill consists entirely of markdown-based guidelines and contains no active code components or external dependencies.
  • [INDIRECT_PROMPT_INJECTION]: The skill defines workflows for processing untrusted external data, such as markdown pipelines and media uploads, which represents an indirect prompt injection attack surface.
  • Ingestion points: SKILL.md identifies markdown pipelines, authored assets, and content models as primary ingestion points.
  • Boundary markers: The instructions do not specify explicit delimiters but advise on using "safe defaults" for content input.
  • Capability inventory: The skill itself defines no capabilities or tools; it only provides guidance for the agent's behavior.
  • Sanitization: The instructions specifically mandate that validation, CSRF safeguards, and permissions must be kept consistent with surrounding admin code.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 09:16 PM
Security Audit — agent-trust-hub — content-management-systems