architect-pro
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized system access were detected. The skill operates purely through text transformation and file generation based on architectural principles.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user project descriptions to generate detailed architectural briefs, creating a common input surface.
- Ingestion points: User-supplied project context (location, lifestyle, functional needs) ingested in SKILL.md.
- Boundary markers: The skill utilizes structured Markdown headers and predefined templates in templates/output-template.md to organize output.
- Capability inventory: The skill writes conceptual reports to local files in Markdown format.
- Sanitization: The skill instructions emphasize adherence to climate constraints and building codes, and include explicit warnings that the output must be validated by a licensed architect.
Audit Metadata