design-system

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were detected. The skill follows a strict human-in-the-loop protocol for all file writes.
  • [NO_CODE]: The skill consists of instructional text and does not include scripts, binaries, or unverifiable dependencies.
  • [PROMPT_INJECTION]: The skill ingests data from local project files (e.g., design/gdd/.md) which presents a theoretical surface for indirect prompt injection. This is mitigated by the collaborative protocol where all drafts are reviewed and approved by the user before writing. Ingestion points: design/gdd/.md. Capability inventory: Read, Write, Edit, Task. Sanitization: Human-in-the-loop approval.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 03:11 AM