typo3-design-system-page
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local commands such as
ddev describe,ddev typo3 cache:flush, andpdfinfoto manage the TYPO3 project and verify generated artifacts. It also runs a bundled Python scriptscripts/audit_design_tokens.pyfor deterministic validation of design tokens. - [INDIRECT_PROMPT_INJECTION]: The skill audits site assets and rendered content, which introduces a potential surface for instructions embedded in external CSS or HTML. To mitigate this, the skill requires the agent to verify all assertions against local source code and use a validation script to confirm findings.
- [DATA_EXFILTRATION]: The skill reads the local
.ddev/config.yamlfile to identify the project root and installation details. This access is localized for project identification, and no evidence of exfiltration to external domains was found.
Audit Metadata