lw-firewall-custom-form-adapter
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as technical documentation for developers to build adapters for the LW Firewall plugin. It provides legitimate PHP code examples for form validation.
- [INDIRECT_PROMPT_INJECTION]: The skill describes an interface for processing untrusted form data. It mitigates potential abuse by recommending strict validation logic, including honeypot checks, signed render tokens, and input sanitization using standard WordPress functions like
sanitize_text_field. - Ingestion points: Untrusted data enters the adapter through the
validate(array $input)method inSKILL.md. - Boundary markers: The code is structured as a class-based utility; no specific prompt boundary markers are required for this implementation code.
- Capability inventory: The code interacts with plugin settings (
Options::get) and storage backends (lw_firewall_resolve_storage). It does not perform network operations, file writes, or shell command execution. - Sanitization: The adapter implements
sanitize_text_fieldfor tokens and enforces a strict empty-string check for honeypot fields.
Audit Metadata