architect
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external codebase context and user requirements to generate architectural designs and implementation code, creating a potential surface where malicious content in processed files could influence agent behavior.
- Ingestion points: Codebase subsystems and system context are ingested during 'Phase A: Ground the problem' and passed as grounding artifacts to design runners.
- Boundary markers: The skill defines a clear multi-phase workflow and discipline for design runners, but does not specify technical delimiters or automated sanitization for the ingested source material.
- Capability inventory: The skill facilitates modification of the local codebase by replacing 'not implemented' placeholders with functional code during Phase D.
- Sanitization: No explicit filtering, validation, or sanitization of source material processed during the grounding phase is mentioned.
Audit Metadata