theme-factory

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [Prompt Injection] (SAFE): The instructions do not contain commands to override safety protocols or ignore system prompts.
  • [Data Exposure & Exfiltration] (SAFE): No credentials or sensitive data access patterns were found.
  • [Indirect Prompt Injection] (LOW): The skill processes untrusted content from user artifacts and custom theme inputs. 1. Ingestion points: Reads artifact content and user-provided theme requirements. 2. Boundary markers: None specified to separate content from instructions. 3. Capability inventory: Modifies document styling and formatting. 4. Sanitization: No sanitization of ingested content is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:29 PM