The Agent Skills Directory

[SAFE]: The skill consists solely of natural language instructions for architectural analysis and contains no executable code, remote downloads, or dangerous command patterns.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection because it is designed to ingest and analyze untrusted external data such as code diffs and architectural designs.
Ingestion points: The skill analyzes code boundaries, diff areas, and component interactions as specified in the 'Return' and 'Focus' sections of SKILL.md.
Boundary markers: No specific delimiters or instructions to ignore embedded commands within the analyzed content are provided.
Capability inventory: The skill is constrained to a read-only sandbox mode and does not possess any subprocess, file-writing, or network capabilities.
Sanitization: There are no mechanisms described for sanitizing or validating the integrity of the analyzed architectural data.

architect-reviewer