wordpress-plugin-development-v2
Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection in the provided code examples for AI integration.
- Ingestion points: The
my_plugin_generate_ai_summaryfunction inSKILL.mdingests untrusted content from the$post->post_contentvariable. - Boundary markers: Absent. The instruction snippet uses direct string concatenation (e.g.,
'Create a compelling 2-sentence summary... ' . substr($content, 0, 1000)) to build AI prompts, providing no isolation between the developer's instructions and the user's data. - Capability inventory: The skill utilizes the
wp_ai_client_promptfunction (identified inSKILL.md), which invokes external AI capabilities through a connector. - Sanitization: The code uses
strip_tags(), which removes HTML formatting but does not sanitize against natural language prompt injection attacks intended to override AI behavior.
Audit Metadata