> ## Documentation Index
> Fetch the complete documentation index at: https://docs.matterai.so/llms.txt
> Use this file to discover all available pages before exploring further.

# Advanced Reviews

> Learn how to use advanced AI-powered code review with tool-use enabled analysis

<Note>
  Advanced AI Code Review is available on all plans.
</Note>

This guide walks you through the advanced AI code review capabilities of MatterAI, which provides deeper, more comprehensive analysis using tool-use enabled AI.

## What is Advanced AI Code Review?

The `/matter review-max` command performs deep, tool-enabled code analysis that goes beyond surface-level scanning. The AI can read multiple files, search your codebase for patterns, and provide comprehensive feedback with higher accuracy.

## Key Capabilities

### Tool-Use Enabled Analysis

* **Multi-file reading**: AI reads relevant files across your codebase
* **Pattern search**: Searches for similar code patterns and implementations
* **Context awareness**: Understands your project's architecture and conventions
* **Dependency analysis**: Analyzes relationships between components

### Deeper Analysis

* **Architectural issues**: Identifies design pattern violations and architectural concerns
* **Security vulnerabilities**: Detects potential security risks and best practice violations
* **Performance bottlenecks**: Finds performance issues and optimization opportunities
* **Code maintainability**: Suggests improvements for long-term code health

### Confidence Scoring

Each review comment includes two confidence scores:

* **Detection confidence**: How confident the AI is about the issue
* **Fix confidence**: How confident the AI is about the proposed solution

### Smart Suggestions

* **Precise code suggestions**: Provides exact code changes with line numbers
* **Context-aware recommendations**: Suggestions consider your project's specific context
* **Best practice alignment**: Recommendations follow industry best practices

## When to Use Advanced AI Code Review

### Best For:

* **Complex PRs with multiple files**: When changes span across many files
* **Security-sensitive changes**: For authentication, authorization, or data handling code
* **Performance-critical code**: For algorithms, database queries, or rendering code
* **Architectural reviews**: When making significant structural changes

### Not Recommended For:

* Simple documentation changes
* Minor formatting or style fixes
* Straightforward bug fixes with clear solutions

## How to Use

1. **Trigger the review**: Comment on any PR with `/matter review-max` or `@matterai review-max`
2. **Wait for analysis**: The AI will analyze your code using tool-use capabilities
3. **Review the feedback**: Examine the comprehensive review comments with confidence scores
4. **Apply suggestions**: Use the provided code suggestions to improve your code

## Example Workflow

<Frame>
  <img className="block" src="https://mintcdn.com/gravitycloud-9ebb5c50/0G_3_GyjEL_0_lmh/images/features/matterai-review-max.png?fit=max&auto=format&n=0G_3_GyjEL_0_lmh&q=85&s=3d9c4463b939c2fd1c477be77c3a1ce2" alt="Advanced AI Code Review Example" width="1894" height="1200" data-path="images/features/matterai-review-max.png" />
</Frame>

## Integration with Other Commands

Advanced AI Code Review works seamlessly with other MatterAI commands:

* **Use with `/matter fix`**: After getting advanced reviews, use the fix command to automatically apply suggestions
* **Combine with `/matter summary`**: Get both deep analysis and high-level overview
* **Follow with `/matter explain`**: Get detailed explanations of complex changes

## Benefits

* **3x more comprehensive**: Tool-use enables deeper analysis than standard reviews
* **Higher accuracy**: Confidence scoring helps prioritize feedback
* **Context-aware**: Understands your specific codebase and patterns
* **Time-efficient**: Get thorough reviews without manual deep-dive analysis

## Tips for Best Results

1. **Provide context**: Include relevant background information in your PR description
2. **Link related issues**: Reference relevant tickets or documentation
3. **Review confidence scores**: Focus on high-confidence suggestions first
4. **Ask follow-up questions**: Use `/matter <question>` to clarify complex suggestions
