BeClaude

revu

New
Community RegistryGeneralby k-park · MIT

Code review plugins for quality, security, architecture, documentation, and commit message assessment

Community PluginView Source

Overview

A curated plugin marketplace for Claude Code, providing specialized AI agents for development productivity.

![License: MIT](https://opensource.org/licenses/MIT) ![Claude Code](https://claude.com/claude-code)

Overview

cc-lib is a collection of production-ready plugins that extend Claude Code's capabilities. Each plugin contains specialized agents designed for specific development tasks - from code reviews and testing to documentation generation and autonomous development loops.

What It Provides

CategoryPluginsAgents
OrchestrationorchTask orchestration, parallel execution, context management
TestingtestTest generation, execution, and failure resolution
Code ReviewrevuCode, architecture, security, documentation, and commit reviews
Feature DevfeatDesign, implementation planning, and scaffolding
DocumentationdocuAPI docs, README generation, code explanation
DebuggingfixCode cleaning and systematic debugging
AutonomousralphRALPH - iterative autonomous development loop

Installation

Add the Marketplace

bash
/plugin marketplace add k-park/cc-lib

Install Plugins

bash
# Orchestration - complex task coordination
/plugin install orch@cc-lib

# Testing - complete test workflow
/plugin install test@cc-lib

# Code Review - multi-dimensional code analysis
/plugin install revu@cc-lib

# Feature Development - from design to implementation
/plugin install feat@cc-lib

# Documentation - generate docs automatically
/plugin install docu@cc-lib

# Debugging - clean and fix code
/plugin install fix@cc-lib

# RALPH - autonomous iterative development
/plugin install ralph@cc-lib

Manage Plugins

bash
/plugin list                    # List installed plugins
/plugin info orch               # Show plugin details
/plugin remove orch@cc-lib      # Remove a plugin

Plugins

orch - Orchestration

Agent orchestration for complex multi-faceted tasks with parallel execution and context management.

AgentDescription
orch:taskComplex task delegation and multi-agent coordination
orch:parallelBreak down tasks into parallel executable subtasks
orch:contextContext management for multi-step workflows

test - Testing

Complete testing workflow - generate tests, run them, and fix failures automatically.

AgentDescription
test:genGenerate comprehensive tests (unit, integration, edge cases)
test:runExecute tests and analyze results with actionable feedback
test:fixFix failing tests by analyzing root causes and implementing corrections

revu - Code Review

Multi-dimensional code review covering quality, security, architecture, and documentation.

AgentDescription
revu:codeCode quality, style, potential bugs, performance, and maintainability
revu:archArchitecture review for design patterns and SOLID principles
revu:secSecurity vulnerabilities and OWASP Top 10 compliance
revu:docDocumentation completeness, clarity, and accuracy
revu:commitCommit message quality and conventional compliance

feat - Feature Development

Feature development agents covering design, implementation, and scaffolding.

AgentDescription
feat:designFeature planning, requirements analysis, and API design
feat:implImplementation planning with code structure and integration points
feat:scaffoldProject scaffolding with file structure and boilerplate

docu - Documentation

Documentation generation for APIs, READMEs, and code explanations.

AgentDescription
docu:genGenerate API documentation from code
docu:readmeCreate comprehensive project README files
docu:explainExplain code logic and architecture

fix - Debugging

Code cleaning and systematic debugging agents.

AgentDescription
fix:cleanRefactor code, remove technical debt, improve maintainability
fix:debugSystematic debugging with root cause analysis

ralph - Autonomous Loop

RALPH (Recursively Adaptive Loop for Progressive Habitability) - autonomous iterative development.

CommandDescription
ralph:goStart autonomous iterative development loop
ralph:stopStop the RALPH loop gracefully with completion summary

Usage Examples

Orchestrate Complex Tasks

bash
# Delegate complex feature to orchestrator
"Build a user authentication system with OAuth2, login UI, and backend APIs"
# Claude automatically delegates to specialized sub-agents in parallel

Complete Testing Workflow

bash
/test:gen        # Generate tests for new feature
/test:run        # Run tests and see results
/test:fix        # Automatically fix any failing tests

Code Review

bash
/revu:code       # Review code quality
/revu:sec        # Check for security issues
/revu:arch       # Evaluate architecture decisions

Feature Development

bash
/feat:design     # Plan the feature architecture
/feat:impl       # Get implementation steps
/feat:scaffold   # Generate project structure

Autonomous Development

bash
/ralph:go        # Let Claude work until completion
/ralph:stop      # Stop and get summary

For Developers

Project Structure

code
cc-lib/
├── agents/              # Source agent definitions
│   ├── debugging/       # Debugging agents
│   ├── documentation/   # Documentation agents
│   ├── feature/         # Feature development agents
│   ├── orchestration/   # Orchestration agents
│   ├── ralph/           # RALPH agents
│   ├── reviews/         # Review agents
│   └── testing/         # Testing agents
│
├── plugins/             # Plugin distribution structure
│   ├── docu/           # Documentation plugin
│   ├── feat/           # Feature development plugin
│   ├── fix/            # Debugging plugin
│   ├── orch/           # Orchestration plugin
│   ├── ralph/          # RALPH plugin
│   ├── revu/           # Code review plugin
│   └── test/           # Testing plugin
│
├── installer/          # Build and installation tools
│   ├── cli/build.sh    # Build script
│   ├── sets/           # Installation presets
│   └── templates/      # Settings templates
│
└── output/.claude/     # Build output (flat structure)

Building

bash
# Run the build script
./installer/cli/build.sh

# Output is generated in output/.claude/

The build process:

  1. Scans agents/ for all .md files
  2. Resolves symbolic links from plugins
  3. Copies files to output/.claude/agents/ (flat structure)
  4. Generates output/.claude/settings.json with absolute paths

Creating Agents

Each agent is a Markdown file with YAML frontmatter:

yaml
---
name: agent-id
description: What this agent does (50-1000 characters)
model: sonnet  # sonnet|haiku|opus
color: blue    # optional
tags: [category, keywords]
examples:
  - context: When to use this agent
    input: Example input
    output: Expected output
    explanation: Why it works
---

Installation Presets

Pre-configured agent sets for different environments:

PresetEnvironmentDescription
minimalPCMinimal set for basic usage
developerPCFull-featured development environment
mobile-basicMobileLightweight agents for mobile
server-ciServerCI/CD pipeline automation

Documentation

License

MIT License - see LICENSE for details.

Contributing

Contributions are welcome! Please see our contributing guidelines for details.


Made with Claude Code

Install & Usage

1
Create the skills directory
mkdir -p .claude/skills
2
Download the skill file
mkdir -p .claude/skills && curl -o .claude/skills/revu.md https://raw.githubusercontent.com/k-park/cc-lib/main/SKILL.md
3
Invoke in Claude Code
/revu
View source on GitHub
securitycode-reviewdocumentationpluginreviewcode-qualityarchitecturecommits

Frequently Asked Questions

What is revu?

Code review plugins for quality, security, architecture, documentation, and commit message assessment

How to install revu?

To install revu, create the .claude/skills directory in your project, then run the curl command to download the skill file. Once installed, invoke it in Claude Code with /revu.

What is revu best for?

revu is a community categorized under General. It is designed for: security, code-review, documentation, plugin, review, code-quality, architecture, commits. Created by k-park.