ln-24-architecture-auditor

Audits implemented architecture boundaries, contracts, dependencies, and configuration ownership against repository evidence.

556|83|Updated Oct 24, 2025
One-click install
npx skills add https://github.com/levnikolaevich/claude-code-skills --skill ln-24-architecture-auditor
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ln-24-architecture-auditor
Source: https://github.com/levnikolaevich/claude-code-skills/tree/main/plugins/codebase-audit-suite/skills/ln-24-architecture-auditor
Command: npx skills add https://github.com/levnikolaevich/claude-code-skills --skill ln-24-architecture-auditor

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Teams often cannot tell whether the architecture their system actually executes still fits current product needs, because documentation drifts, hidden cycles accumulate, and cross-component ownership becomes ambiguous. This Skill performs a read-only, evidence-driven audit of the real architecture so structural risks are found before they become correctness, deployment, or maintenance failures.

Core Features & Use Cases

  • Actual Architecture Discovery: Maps modules, layers, entrypoints, deployment units, and runtime wiring from executable evidence rather than folder names or stale diagrams.
  • Contract and Dependency Analysis: Traces dependency direction, cycles, forbidden imports, event schema agreement, and configuration ownership with materiality gating.
  • Verdict-Based Reporting: Produces a prioritized P0-P3 findings table with PASS, CONCERNS, FAIL, or BLOCKED verdicts and verified practice references.
  • Use Case: Before a major refactor, run the audit to confirm whether a suspected cross-layer dependency actually creates change amplification, and get the smallest safe boundary correction with migration risk assessed.

Quick Start

Ask the agent to audit this repository's architecture for boundary violations, dependency cycles, and ownership ambiguity, and return a verdict with prioritized findings.

Frequently Asked Questions about ln-24-architecture-auditor

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I audit software architecture in an existing codebase?

Run an evidence-based architecture audit that maps modules, entrypoints, and runtime wiring from executable code rather than documentation. This Skill traces critical flows, checks dependency direction and contracts, then reports prioritized findings with a PASS, CONCERNS, FAIL, or BLOCKED verdict.

How to detect dependency cycles and layer violations in code?

Build the module dependency graph from resolved internal edges, accounting for aliases, re-exports, generated code, and runtime loading. A cycle or cross-layer call only counts as a finding when it creates a concrete change, testing, deployment, or failure cost.

Does an architecture audit modify my code?

No, the audit is strictly read-only. It inspects files, manifests, git history, and runtime registration without modifying code or architecture documents, and discloses any permitted diagnostic caches it uses.

When should I use an architecture audit instead of a code review?

Use an architecture audit for system-level structure: boundaries, ownership, contracts, and dependency topology. Use code review for local correctness, and a persistence-focused review for query, transaction, and data-resource correctness, which this audit deliberately excludes.

What are the limitations of a static architecture audit?

Components wired only at runtime may be undiscoverable statically, so the audit verifies registration paths and marks unverifiable wiring as BLOCKED rather than guessing. Inferred boundary rules are reported with confidence levels instead of being presented as policy.