What standards does this skill validate against?

The skill validates code against lev-ref patterns, specifically checking for @lev-os namespace usage, package.json leviathan fields, and fractal directory compliance.

Where are the audit reports stored?

Reports are saved in XDG-compatible scratch folders, typically located at ~/.config/lev/scratch/code-review/ categorized by session timestamps.

How does the parallel inspection work?

The skill launches multiple agents simultaneously in phases: Phase 1 focuses on structure and configuration, Phase 2 handles deep analysis of imports and tests, and Phase 3 synthesizes all findings into a final report.

What is lev-code-review?

It is a multi-agent orchestrator for Claude Code that performs parallel inspections of codebase structure, modularity, and compliance with specific architectural patterns.

Can I resume a long-running audit?

Yes, the skill supports a --continue flag that allows you to resume analysis from a specific session directory if the audit was interrupted.

Lev Code Review Orchestrator

Name: Lev Code Review Orchestrator
Author: lev-os

bylev-os

0•

コラボレーションツール

Conducts deep code organization and modularity audits using parallel multi-agent inspection and XDG-compatible reporting.

Lev Code Review is a specialized multi-agent orchestrator designed to automate comprehensive codebase audits. It dispatches parallel inspection agents to validate directory structures, package namespaces, and fractal compliance against lev-ref standards. By generating XDG-compatible markdown reports across three distinct phases—Structure Analysis, Deep Analysis, and Synthesis—it provides a systematic way to ensure architectural consistency, identify structural anomalies, and verify dependency health in modular software projects.

主な機能

01Parallel multi-agent execution across three specialized analysis phases

02XDG-compatible session management and markdown artifact generation

03Tiered validation gates for both universal and module-specific checks

04Confidence-scored findings with actionable remediation recommendations

05Automated validation of @lev-os namespace and fractal compliance patterns

060 GitHub stars

ユースケース

01Detecting orphaned modules, nested duplicates, and dependency hoisting issues

02Enforcing architectural standards and fractal compliance in core packages

03Auditing large-scale modular repositories for structural and naming consistency

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add lev-os/agents lev-code-review

For use in Claude.ai and ChatGPT

Download Skill