magento2-backend-toolkit
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input for module names, labels, and descriptions which are then interpolated into generated PHP and XML files.\n- Ingestion points: User responses to mandatory questions for all features (e.g., Module name, Parameters, Labels) as defined in
SKILL.md.\n- Boundary markers: The skill includes instructions to use quoted strings for PHP and standard XML escaping, along with Magento's Escaper class for HTML/PHTML output.\n- Capability inventory: The generated files include PHP classes, XML configuration, and PHTML templates; no direct exec() or network operations are included in the templates.\n- Sanitization: The skill mandates regex validation for identifiers (e.g., [A-Z][A-Za-z0-9_]* for module names) and explicitly refuses inputs containing PHP tags or punctuation.\n- [DYNAMIC_EXECUTION]: The skill generates executable PHP scripts and configuration files (XML/JSON) based on pre-defined templates in the assets/templates/ directory.\n- Evidence: Templates such as assets/templates/plugin.php.tpl and assets/templates/email_helper.php.tpl are used to assemble new code files at runtime based on user configuration.
Audit Metadata