design-system-management
Pass
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting data from external design software (Figma, Sketch, Penpot) and local configuration files to generate source code in various formats (Swift, CSS, JavaScript). Maliciously crafted input in the source design files could potentially influence the generated output. \n- Ingestion points: Data is ingested from
design-tokens.jsonand external design files as described in the Synchronization Workflow section. \n- Boundary markers: The skill includes a manual confirmation step ('Apply changes with confirmation') and maintains change history in a.tokens-history/directory. \n- Capability inventory: The skill uses the/sync-design-systemcommand to write files to the filesystem and generate code assets. \n- Sanitization: The skill documents 'Validation Rules' including checks for type consistency, naming compliance, and reference integrity to verify token data before application.
Audit Metadata