typo3-backend-rights

Pass

Audited by Gen Agent Trust Hub on Oct 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides instructions and bundled PHP scripts (audit-backend-rights.php, apply-backend-rights.php, apply-page-permissions.php) designed to be executed via ddev exec. These scripts perform write operations on the TYPO3 database, including the be_groups, be_users, and pages tables, which allows for substantial modification of system access controls.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the TYPO3 database (such as page titles, group titles, and content element types) to generate audit reports and plans. This creates a surface for indirect prompt injection if an attacker has already placed malicious instructions within the TYPO3 records.
  • Ingestion points: Reads from be_groups, be_users, pages, tt_content, and sys_filemounts database tables across all audit and application scripts.
  • Boundary markers: The scripts output structured JSON and rely on explicit AI instructions to classify data, but do not employ specific prompt delimiters for data values.
  • Capability inventory: The skill can perform database inserts and updates via Doctrine\DBAL\Connection and TYPO3's DataHandler in apply-backend-rights.php and apply-page-permissions.php.
  • Sanitization: PHP scripts perform validation on inputs, such as character length checks on group titles and existence checks for page and user UIDs before performing transactions.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 14, 2026, 07:35 PM