backend-architect
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWNO_CODE
Full Analysis
- [No Code] (SAFE): The skill consists entirely of markdown instructions. No scripts, binaries, or configuration files for automated execution were found.
- [Indirect Prompt Injection] (LOW): This skill has an attack surface as it processes untrusted user requirements. 1. Ingestion points: User input for system design. 2. Boundary markers: Absent. 3. Capability inventory: No tool calls, file writes, or network operations defined. 4. Sanitization: Absent. The lack of functional capabilities limits this to a low risk.
- [Data Exposure] (SAFE): No patterns suggesting access to sensitive local files or hardcoded credentials were identified.
- [External Downloads] (SAFE): No remote resources or package managers are invoked.
Audit Metadata