wordpress-content
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to publish user-provided content verbatim to a WordPress site. This preserves the original formatting and prevents the agent from interpreting instructions within the content, but it inherently allows untrusted data to be sent to the remote WordPress server.\n
- Ingestion points: User-supplied body content processed as defined in
SKILL.mdandreferences/verbatim-publishing.md.\n - Boundary markers: Content is handled through structured JSON payloads and temporary files, with explicit instructions to avoid rewriting or interpreting the content.\n
- Capability inventory: The skill possesses network write capabilities (POST/PUT/PATCH/DELETE) to user-defined WordPress endpoints via
curland local file system access for configuration storage.\n - Sanitization: The skill deliberately avoids sanitizing input to maintain fidelity, relying on the target WordPress server's built-in security filters (such as KSES) and REST API validation.
Audit Metadata