A sophisticated multi-agent workflow that automates the process of writing code and performing peer reviews through specialized AI roles.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install code-review-cycle-skill
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install code-review-cycle-skill using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Code Review Cycle?
The Code Review Cycle is a technical framework designed for the Openclaw Skills ecosystem to simulate a professional development environment. It orchestrates a collaborative loop between two distinct agents: a Coder (Agent A) and a Reviewer (Agent B). While the Coder focuses on implementing features and modifying files, the Reviewer maintains a strict read-only perspective to audit the changes, identify bugs, and suggest architectural improvements. This separation of concerns ensures that code is not just written, but validated through an iterative feedback loop before final approval.
Code Review Cycle Use Cases
- Implementing complex features where a second set of AI eyes is needed to catch edge cases.
- Refactoring legacy modules with an automated audit to prevent regressions.
- Enforcing high coding standards by using a specialized reviewer agent to critique implementation logic.
- Creating autonomous self-correction loops where the AI iterates on a solution until it passes peer review criteria.
How Code Review Cycle Works
- The orchestrator spawns Agent A to implement the user's request, generating code diffs and implementation summaries.
- The system then spawns Agent B in a read-only environment to review the changes made by Agent A.
- Agent B outputs a structured report identifying critical issues, optimization suggestions, and a merge conclusion.
- If the --rounds parameter is set and issues are found, the system automatically feeds the feedback back to Agent A for a new iteration.
- The process continues until the specified rounds are exhausted or the Reviewer approves the code, providing a full history in the main session.
Code Review Cycle Setup
To use this within your Openclaw Skills enabled environment, you can trigger the cycle directly from the command interface:
# Basic usage
/cr "Implement user login validation"
# Advanced usage with specific agents and 2 automated rounds
/cr --agent-a codex --agent-b claude-code --rounds 2 "Refactor utils/date.ts"
Code Review Cycle Data Schema & Taxonomy
| Role | Responsibility | Permissions |
|---|---|---|
| Agent A (Coder) | Feature implementation and file modification | Read/Write |
| Agent B (Reviewer) | Logical audit and quality feedback | Read-Only |
| Main Session | Orchestration, context passing, and history tracking | Controller |
| Output Diffs | Standardized markdown summaries of code changes | Metadata |
Code Review Cycle Advanced Features
- Multi-agent orchestration supporting various models like codex and claude-code for different roles.
- Automated iteration cycles using the rounds parameter to enable autonomous improvement.
- Role-based access control ensuring the Reviewer agent cannot accidentally modify the codebase.
- Ephemeral session management where agent contexts are cleared after each round to maintain focus and prevent context bloat.
- Support for custom working directories to target specific microservices or modules.
SKILL.md
Loading
Related Openclaw Skills
Volume-Price Analysis AI Skill
An expert technical analysis skill that validates the authenticity of price trends in the A-share market by analyzing trading volume patterns.
Fundamental Screening
A sophisticated financial analysis tool designed to filter A-share stocks based on growth, value, and risk mitigation metrics.
Watch for Drive Changes
A specialized recipe for establishing real-time event subscriptions on Google Drive files and folders.
AI Employee Configuration Service
A specialized service to configure and deploy OpenClaw AI agents tailored for automated business operations and organizational efficiency.
AI Image Generation Assistant
A sophisticated AI-powered toolkit for generating, editing, and batch-processing high-quality visual assets from text or reference images.
opc-contract-manager
A comprehensive legal operations copilot designed to help solo entrepreneurs review contracts, mitigate risks, and manage document lifecycles without a dedicated legal team.
Featured*
