architecture-guardlisted

# Architecture Guard Skill ## Goal This skill protects architectural integrity through: - **Automated Audits**: Script-based violation detection - **Manual Reviews**: Deep architectural analysis across multiple dimensions ## When to Use Activate this skill when: - Planning new features (understand architectural impact) - Performing code reviews (check architectural consistency) - Refactoring core modules - Comparing multiple design approaches ## How to Use ### Layer 1: Automated Audit Run the automated audit script to detect rule violations: ```bash node .agent/skills/architecture-guard/scripts/index.js ``` The script will: 1. Read architectural rules from `.agent/rules/architecture-design.md` 2. Scan source files for violations 3. Output structured violation reports **Analyze Output**: Present findings to the user with: - Clear explanation of each violation - Why it's problematic - Suggested solutions ### Optional: project-specific audit script Some teams add an **extra** Node entry (e.g. `.agent/skills/architecture-guard/scripts/custom-audit.cjs` or a sibling `architecture-audit/scripts/index.cjs`) that encodes **repository-specific** checks (import boundaries, forbidden paths, etc.). If such a script exists and is safe to run: 1. Run it after (or instead of) the default `index.js` when the user asks for a stricter or custom gate. 2. Merge its output with Layer 1 results; deduplicate and prioritize by severity. 3. If **no** script is present, rely on `index.js` wh