svg-logo-creator
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Prompt Injection] (SAFE): No patterns of instruction override, jailbreak attempts, or system prompt extraction were found in the skill text.
- [Data Exposure & Exfiltration] (SAFE): The skill does not contain hardcoded credentials or sensitive file paths, and it lacks any network-accessing capabilities.
- [Unverifiable Dependencies & Remote Code Execution] (SAFE): There are no requests for external packages (npm/pip) or instructions to download and execute remote scripts.
- [Indirect Prompt Injection] (SAFE): 1. Ingestion points: The skill processes user-provided logo concept briefs and descriptions. 2. Boundary markers: No explicit delimiters or warnings are used for external data. 3. Capability inventory: The skill has no associated tools, subprocess calls, or file-writing permissions. 4. Sanitization: No sanitization is implemented, but the lack of capabilities renders the injection surface non-exploitable.
Audit Metadata