coderabbit-fix-flow

# CodeRabbit Fix Flow

## Overview

This skill automates the workflow of processing CodeRabbit code review feedback by saving the review output to a timestamped document, then using MCP tools (sequential thinking and Exa context) to analyze and implement fixes with minimal code changes.

## When to Use

Use this skill immediately after running `coderabbit --plain` or when you have CodeRabbit feedback that needs systematic processing. The skill handles type safety issues, code style violations, and other CodeRabbit-identified problems.

## Workflow

### Step 1: Execute CodeRabbit Review

Run the CodeRabbit review command in plain text mode:
```bash
coderabbit --plain
```

### Step 2: Save Feedback Document

Save the CodeRabbit output to a timestamped QA document:
- Create file: `memory-bank/qa/coderabbit/cr-qa-{timestamp}.md`
- Include the full CodeRabbit output in the document
- Add YAML front matter with metadata:
  ```yaml
  ---
  title: "CodeRabbit QA Review - {timestamp}"
  link: "cr-qa-{timestamp}"
  type: "qa"
  tags:
    - code-review
    - coderabbit
    - type-safety
  created_at: "{timestamp}"
  updated_at: "{timestamp}"
  uuid: "{generate-uuid}"
  ---
  ```

### Step 3: Analyze Issues with Sequential Thinking

Use the sequential thinking MCP tool to analyze all identified issues:
1. **Categorize issues** by type (type safety, performance, style, security)
2. **Prioritize fixes** (critical runtime issues first, then documentation)
3. **Plan minimal changes** to achieve the fixes
4. **Identify dependencies** between issues

### Step 4: Get Best Practices Context

Use the Exa code context MCP tool to research current best practices for each issue type:
- For type issues: "TypeScript type guards runtime validation best practices"
- For Python type issues: "Python type annotations Optional None best practices"
- For performance: "Performance optimization best practices [language]"
- For security: "Security vulnerability fixes [language]"

### Step 5: Implement