architecture-designer
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides architectural patterns, ADR templates, and non-functional requirement checklists in markdown format. No malicious patterns, such as prompt injection, obfuscation, or data exfiltration, were detected. The metadata and content are consistent with a standard architectural design assistant.\n- [INDIRECT_PROMPT_INJECTION]: The skill identifies a data ingestion surface where user requirements are processed to create system designs. While the ingestion of untrusted data is a known vector, the skill's lack of dangerous capabilities mitigates the risk.\n
- Ingestion points: Requirement gathering workflow in SKILL.md.\n
- Boundary markers: The instructions do not define specific delimiters to separate untrusted user requirements from the agent's core instructions.\n
- Capability inventory: The skill does not possess capabilities for command execution, subprocess spawning, file system access, or network operations across any of its files.\n
- Sanitization: External requirements are not sanitized or filtered before being processed into architectural documentation.
Audit Metadata