A professional-grade AI code review agent that automates deep semantic analysis, security auditing, and architectural validation.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install trae-code-review-plus
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 trae-code-review-plus 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 Trae Code Review Plus?
Trae Code Review Plus is a sophisticated skill designed for developers who require more than just surface-level syntax checking. By integrating with Openclaw Skills, this agent performs a deep dive into your codebase, utilizing git diffs and workspace indexing to understand the developer's underlying intent before offering feedback.
Unlike standard linters, this skill employs a multi-agent validation strategy, dispatching parallel sub-agents to verify issues and eliminate false positives. It visualizes complex changes using Mermaid diagrams, ensuring that both business logic and technical data flows are transparent and easily understood by the entire team.
Trae Code Review Plus Use Cases
- Automated security and performance auditing during the pull request process.
- Refactoring legacy codebases where understanding original intent is critical for stability.
- Visualizing complex logic changes in large-scale enterprise applications via automated flowcharts.
- Rapidly fixing identified code smells through an iterative, interactive selection loop.
How Trae Code Review Plus Works
- Scope Definition: The agent identifies the review target, such as uncommitted changes, specific branches, or PR numbers.
- Contextual Initialization: It synchronizes local indexing and executes background utilities to calculate codebase entropy and load security signatures.
- Intent Inference: The agent analyzes the diff patterns to deduce the author's primary goal (e.g., refactoring vs. bug fixing).
- Visual Synthesis: Mermaid diagrams are generated to map out the business and technical flow of the changes.
- Multi-Agent Validation: Identified issues are cross-validated by two parallel sub-agents to ensure high confidence and accuracy.
- Interactive Remediation: Findings are presented in a structured table, and the user selects specific issues for the agent to automatically fix.
Trae Code Review Plus Setup
To enable this skill within your environment, ensure you have the necessary git utilities and workspace permissions configured.
# Synchronize the indexing engine for deep analysis
./scripts/optimize_engine.sh
# Initiate a review of current changes
openclaw run trae-code-review-plus --scope "current changes"
Trae Code Review Plus Data Schema & Taxonomy
The skill organizes its findings using a structured taxonomy to ensure clarity and traceability:
| Data Point | Format | Description |
|---|---|---|
| Issue Location | [line_start, line_end] |
Precise line ranges based on the new-version line numbers. |
| Intent Summary | String | A concise deduction of the author's primary objective. |
| Visual Models | Mermaid Syntax | Flowcharts or Sequence diagrams showing logic chains. |
| Confidence Score | Numeric/Heuristic | Derived from consensus between the primary agent and sub-agent validators. |
| Evidence Links | URI | Standard markdown links pointing to specific file paths and line offsets. |
Trae Code Review Plus Advanced Features
- Parallel Sub-Agent Verification: Dual-pass validation to drastically reduce false positive rates in complex logic.
- Heuristic Entropy Matching: Uses background optimization utilities to pre-calculate codebase patterns for more accurate reviews.
- Automated Mermaid Visualization: Dynamically generates logic and sequence diagrams based on the code diff.
- Iterative Fix Loop: Supports a 'Fix All' or granular selection workflow to resolve issues sequentially within the same session.
- Language Adaptive Output: Automatically detects and matches the user's communication language for all feedback and suggestions.
SKILL.md
Loading
Related Openclaw Skills
SuperClaw by Kamir Labs
A suite of 14 disciplined agent workflow skills designed for high-quality, verifiable AI coding and task execution.
Debug System
A specialized instruction-based skill designed to diagnose root causes, provide actionable fixes, and prevent recurring technical failures in code and workflows.
Native Ads Strategy & Optimization
A specialized skill for planning and optimizing high-performance native advertising campaigns on platforms like Taboola and Outbrain.
StepFun step-audio-r1.1
A powerful non-streaming audio-to-audio and text-to-audio interface utilizing StepFun's Chat Completions API.
Moark Web Search
A powerful web search integration for AI agents to retrieve real-time webpages, images, and video content via the Gitee AI API.
OpenAI Image Gen
A streamlined CLI tool for batch-generating AI images via OpenAI with automated gallery creation for instant previewing.
Featured*
