prototype-design-xml
Pass
Audited by Gen Agent Trust Hub on Jul 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or persistence mechanisms were detected in the skill's instructions or template files.
- [DATA_EXPOSURE]: The skill reads user-provided requirements and writes generated prototypes to the local filesystem. This behavior is restricted to its intended design purpose and does not involve network communication or access to sensitive user credentials like SSH or AWS keys.
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests untrusted data (user-provided requirements) to generate output. However, its functionality is limited to producing XML structures, and it lacks dangerous capabilities like shell command execution or network access that could be exploited via injection.
- Ingestion points: User-provided requirement documents (referenced in
SKILL.mdStep 1). - Boundary markers: None identified in the prompt templates.
- Capability inventory: Local file-writing to the
output/prototype/drawiodirectory. - Sanitization: None identified for the input requirement documents.
Audit Metadata