creating-themes

Pass

Audited by Gen Agent Trust Hub on Aug 31, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to 'replicate liberated sites', which involves ingesting and processing data from external websites (HTML, CSS, fonts, logos) to generate WordPress theme files, code, and configurations.
  • Ingestion points: Data from external sites enters the agent's context during the replication process (specifically mentioned in the 'Fonts' and 'Header (replica)' sections of SKILL.md).
  • Boundary markers: There are no explicit instructions or delimiters to isolate untrusted data from the source site to prevent it from influencing the agent's behavior during theme generation.
  • Capability inventory: The skill instructs the agent to create a complex file structure, write PHP code (functions.php), generate JSON configuration (theme.json), and perform network downloads for assets like fonts and logos.
  • Sanitization: While the skill mentions sanitizing specific values like line-heights, it lacks a comprehensive mechanism to filter or escape instructions that might be embedded in the source site's metadata, CSS comments, or text content.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 31, 2026, 06:40 AM
Security Audit — agent-trust-hub — creating-themes