https://github.com/cskwork/llm-review
Claude (planner + coder) and codex (reviewer)
https://github.com/cskwork/llm-review
Last synced: 21 days ago
JSON representation
Claude (planner + coder) and codex (reviewer)
- Host: GitHub
- URL: https://github.com/cskwork/llm-review
- Owner: cskwork
- License: gpl-3.0
- Created: 2026-01-16T21:13:03.000Z (5 months ago)
- Default Branch: main
- Last Pushed: 2026-01-17T00:42:36.000Z (5 months ago)
- Last Synced: 2026-01-17T11:18:02.445Z (5 months ago)
- Language: Shell
- Size: 160 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Claude Codex Marketplace
A Claude Code plugin marketplace providing multi-AI orchestration tools for planning, implementing, and reviewing code changes.
**Supported Platforms:** Windows, macOS, Linux
> **New to Claude Codex?** Check out our [comprehensive Wiki](https://github.com/cskwork/claude-codex/wiki) for beginner-friendly guides and tutorials.
## Why Multi-AI Review?
**Would you deploy code that only one person looked at?**
In professional software development, code reviews are mandatory. Google requires review for every change. Microsoft uses multiple review stages. Claude Codex brings this professional standard to AI-assisted development.
Instead of trusting a single AI's output, your code goes through **two independent reviews**:
| Reviewer | What It Catches |
|----------|-----------------|
| **Claude Sonnet** | Obvious bugs, security basics, code style |
| **Codex** | Fresh perspective from a different AI |
**The result:** More bugs caught, better security, and production-ready code.
Each reviewer checks for OWASP Top 10 vulnerabilities, proper error handling, and code quality. If Sonnet misses something, Codex will likely catch it. The loop-until-approved model means code doesn't proceed until both reviewers give the green light.
Learn more: [Why Claude Codex?](https://github.com/cskwork/claude-codex/wiki/Why-Claude-Codex)
## Installation
### Step 1: Add Marketplace
```bash
/plugin marketplace add cskwork/claude-codex
```
### Step 2: Install Plugin
```bash
# Install at user scope (available in all projects) - RECOMMENDED
/plugin install claude-codex@claude-codex --scope user
# OR install at project scope (this project only)
/plugin install claude-codex@claude-codex --scope project
```
### Step 3: Add .task to .gitignore
The plugin automatically creates `.task/` in your project directory. Add it to `.gitignore`:
```bash
echo ".task" >> .gitignore
```
> **Multi-Project Support:** When installed at user scope, you can run the pipeline in multiple projects simultaneously. Each project gets its own isolated `.task/` directory.
### Usage
After installation, use skills with the plugin namespace:
```bash
# Start the full pipeline
/claude-codex:multi-ai Add user authentication with JWT
# Or run individual skills
/claude-codex:review-sonnet
/claude-codex:review-codex
/claude-codex:implement-sonnet
```
## Available Plugins
### claude-codex
Multi-AI orchestration pipeline with sequential review workflow (sonnet → codex).
**Skills included:**
| Skill | Model | Purpose |
| ------------------ | ----------------- | ------------------------------------------- |
| `multi-ai` | - | Pipeline entry point (starts full workflow) |
| `implement-sonnet` | Claude Sonnet 4.5 | Code implementation with main context |
| `review-sonnet` | Claude Sonnet 4.5 | Fast review (code + security + tests) |
| `review-codex` | Codex CLI | Final review via OpenAI Codex |
## Recommended Subscriptions
| Service | Subscription | Purpose |
| --------------- | ------------ | ---------------------------------------------- |
| **Claude Code** | MAX 20 | Main thread (planning, coding) + Review skills |
| **Codex CLI** | Plus | Final reviews (invoked via skill) |
## How It Works
### Quick Start with `/multi-ai`
```bash
/claude-codex:multi-ai Add user authentication with JWT tokens
```
This command:
1. Cleans up previous task files
2. Creates and refines a plan
3. Runs sequential reviews (sonnet → codex)
4. Implements the code
5. Runs sequential reviews (sonnet → codex)
6. Marks complete
### Sequential Review Flow
Reviews run **sequentially** - each model reviews only ONCE per cycle:
```
Plan/Code → review-sonnet → fix → review-codex → fix (restart)
```
**Key benefits:**
- Each model provides unique perspective without re-reviewing
- Progressive refinement (fast → final)
- Token-efficient (forked context isolation)
## Marketplace Structure
```
claude-codex/
├── .claude-plugin/
│ └── marketplace.json # Marketplace catalog
├── plugins/
│ └── claude-codex/ # Multi-AI plugin
│ ├── .claude-plugin/
│ │ └── plugin.json # Plugin manifest
│ ├── skills/ # Pipeline skills
│ │ ├── multi-ai/
│ │ ├── implement-sonnet/
│ │ ├── review-sonnet/
│ │ └── review-codex/
│ ├── scripts/ # Orchestration scripts
│ ├── docs/ # Standards and workflow
│ ├── .task.template/ # Task directory template
│ ├── pipeline.config.json
│ ├── CLAUDE.md
│ └── AGENTS.md
└── README.md
```
## Prerequisites
- [Claude Code](https://claude.ai/code) installed and authenticated
- [Codex CLI](https://github.com/openai/codex) installed and authenticated (for review-codex skill)
- [Bun](https://bun.sh/) installed (required by Claude Code, also used for cross-platform JSON processing)
> **Note:** This plugin works on Windows, macOS, and Linux. All shell scripts use Bun for JSON processing instead of `jq`, ensuring cross-platform compatibility out of the box.
## Configuration
The plugin includes `pipeline.config.json` with these settings:
| Setting | Description | Default |
| ----------------------------------- | --------------------------------------------- | ----------------- |
| `autonomy.mode` | `autonomous`, `semi-autonomous`, `supervised` | `semi-autonomous` |
| `autonomy.planReviewLoopLimit` | Max plan review iterations | `10` |
| `autonomy.codeReviewLoopLimit` | Max code review iterations | `15` |
| `errorHandling.autoResolveAttempts` | Retries before pausing | `3` |
### Per-Project Overrides
Create `pipeline.config.local.json` in your project directory to override settings:
```json
{
"autonomy": {
"codeReviewLoopLimit": 20
}
}
```
Config priority: project-local > plugin-local > plugin-base
## Creating Your Own Plugin
To add a new plugin to this marketplace:
1. Create a directory under `plugins/your-plugin-name/`
2. Add `.claude-plugin/plugin.json` with your plugin manifest
3. Add your skills under `skills/`
4. Update `.claude-plugin/marketplace.json` to include your plugin
Example plugin.json:
```json
{
"name": "your-plugin-name",
"version": "1.0.4",
"description": "What your plugin does",
"author": { "name": "Your Name" },
"skills": "./skills/"
}
```
## Troubleshooting
### Plugin not found
```bash
# Verify marketplace is added
/plugin marketplace list
# Re-add if needed
/plugin marketplace add cskwork/claude-codex
```
### Skills not loading
```bash
# Check plugin is installed
/plugin list
# Reinstall if needed
/plugin uninstall claude-codex@claude-codex
/plugin install claude-codex@claude-codex --scope user
```
### Validate plugin structure
```bash
/plugin validate .
```
## Documentation
For detailed guides, visit our [Wiki](https://github.com/cskwork/claude-codex/wiki):
- [Why Claude Codex?](https://github.com/cskwork/claude-codex/wiki/Why-Claude-Codex) - Benefits of multi-AI review
- [Getting Started](https://github.com/cskwork/claude-codex/wiki/Getting-Started) - Step-by-step installation for beginners
- [Understanding the Pipeline](https://github.com/cskwork/claude-codex/wiki/Understanding-the-Pipeline) - How the review process works
- [How to Use](https://github.com/cskwork/claude-codex/wiki/How-to-Use) - Commands and usage examples
- [Configuration](https://github.com/cskwork/claude-codex/wiki/Configuration) - Customization options
- [Troubleshooting](https://github.com/cskwork/claude-codex/wiki/Troubleshooting) - Common issues and solutions
## Related Projects
- [claude-codex-gemini](https://github.com/Z-M-Huang/claude-codex-gemini) - Adds Gemini as a dedicated orchestrator
## License
GPL-3.0 license