architecture-lead

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a strictly defined consultation protocol for architectural decisions, with explicit boundaries for autonomous action and clear escalation paths for novel or risky changes.
  • [DATA_EXPOSURE]: The skill is designed to read local project files and documentation (e.g., tech-stack.md, CLAUDE.md, and src/ directories) to gather context. This behavior is intrinsic to its stated purpose as an architecture advisor and does not involve exfiltration to external domains.
  • [INDIRECT_PROMPT_INJECTION]: The skill evaluates content from the local codebase which could theoretically contain malicious instructions.
  • Ingestion points: Reads files from src/components/, src/lib/, src/app/api/, conductor/tech-stack.md, and CLAUDE.md.
  • Boundary markers: The instructions do not define specific delimiters or warnings to ignore embedded instructions within the project files it analyzes.
  • Capability inventory: The skill generates structured architectural advice in JSON format and can suggest additive schema changes via the Supabase MCP tool; it does not possess arbitrary shell or code execution capabilities.
  • Sanitization: There is no mention of sanitizing or escaping the content retrieved from the codebase before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 07:44 AM
Security Audit — agent-trust-hub — architecture-lead