wordpress-writer-showcase
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches the WordPress Brand Writing Style Guide from the official WordPress GitHub repository. This is an expected vendor-owned resource for marketing guidelines.
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from user-provided 'Site URLs' to generate descriptions, which introduces an indirect prompt injection surface.
- Ingestion points: External site content is retrieved and processed based on the user-supplied
Site URLfield. - Boundary markers: Absent; the instructions do not specifically tell the model to ignore instructions embedded in the external site content.
- Capability inventory: Limited to generating two-paragraph text descriptions for human review. No shell access, file-system writes, or network exfiltration capabilities are utilized.
- Sanitization: The skill relies on manual human review of the generated drafts and the agent's internal safety filters.
Audit Metadata