Code Turtle 🐢 - Azure OpenAI Code Review Proxy

A lightweight Node.js middleware that acts as a bridge between GitHub Actions (or other CI/CD tools) and Azure OpenAI. It provides a specialized endpoint to perform automated code reviews on git diffs using a "Senior Software Engineer" persona.

🚀 Features

• Azure OpenAI Integration: Proxies requests to your secure Azure OpenAI deployment.
• Automated Code Review: Dedicated /review endpoint that accepts a code diff and returns constructive feedback.
• OpenAI Compatibility: Includes a /v1/chat/completions endpoint compatible with standard OpenAI client libraries.
• Secure: Keeps your Azure credentials on the server, exposing only the proxy endpoints.

🛠️ Prerequisites

• Node.js (v18 or higher)
• An Azure OpenAI resource with a deployed model (e.g., GPT-4 or GPT-3.5-Turbo).

📦 Installation

1. Clone the repository:

   git clone https://github.com/spotmies/code_turtle.git
   cd code_turtle

2. Install dependencies:

   npm install

3. Configure Environment Variables: Copy the example environment file and fill in your Azure details.

   cp .env.example .env

   Edit .env:

   AZURE_ENDPOINT=https://your-resource-name.openai.azure.com
   AZURE_API_KEY=your_azure_api_key
   AZURE_DEPLOYMENT=your_deployment_name
   AZURE_API_VERSION=2023-05-15
   PORT=3000

🏃‍♂️ Running Locally

Start the development server with hot-reload:

npm run dev

The server will start at http://localhost:3000.

🔌 API Endpoints

1. Code Review (POST /review)

Submit a git diff for review.

Request:

POST /review
Content-Type: application/json

{
  "diff": "diff --git a/server.js b/server.js..."
}

Response: Returns the AI's critique of the code changes.

2. Chat Completions (POST /v1/chat/completions)

Standard OpenAI-compatible endpoint.

Request:

POST /v1/chat/completions
Content-Type: application/json

{
  "messages": [{ "role": "user", "content": "Hello!" }]
}

🤖 GitHub Actions Integration

1. Add the Workflow to Your Repository

To enable Code Turtle on a repository, copy the workflow file to your target repository:

1. Download or copy the workflow file from this repository: .github/workflows/codeturtle.yml

2. Place it in your target repository at .github/workflows/codeturtle.yml

📄 Click to view the workflow file contents

name: Code Turtle AI Review

on:
  pull_request_target:
    types: [opened, synchronize]

permissions:
  contents: read
  pull-requests: write
  issues: write

jobs:
  ai_review:
    runs-on: ubuntu-latest

    steps:
      - name: Checkout base repository
        uses: actions/checkout@v4
        with:
          ref: ${{ github.event.pull_request.base.sha }}
          fetch-depth: 0

      - name: Fetch PR head
        run: |
          PR_NUMBER=${{ github.event.pull_request.number }}
          git fetch origin refs/pull/$PR_NUMBER/head:pr-head

      - name: Generate PR diff
        run: |
          BASE_SHA=${{ github.event.pull_request.base.sha }}
          git diff "$BASE_SHA"...pr-head > diff.txt
          wc -l diff.txt

      - name: Chunk diff
        run: |
          mkdir -p chunks
          split -l 400 diff.txt chunks/diff_

      - name: Send chunks to Code Turtle
        env:
          API_URL: ${{ secrets.CODE_TURTLE_URL }}/review
        run: |
          # Create prompt file
          cat <<'PROMPT' > prompt.txt
          You are a senior software engineer performing a strict code review.
          Focus on bugs, security, performance, and maintainability.
          Output valid GitHub-flavored Markdown.
          PROMPT

          echo "## 🐢 Code Turtle Review" > combined_review.md

          for CHUNK in chunks/*; do
            RESPONSE=$(curl -s -X POST "$API_URL" \
              -H "Content-Type: application/json" \
              -d "$(jq -n \
                --arg diff "$(cat "$CHUNK")" \
                --arg prompt "$(cat prompt.txt)" \
                '{diff: $diff, prompt: $prompt}')")

            echo "$RESPONSE" | jq -r '.choices[0].message.content // "Review failed."' >> combined_review.md
            echo -e "\n---\n" >> combined_review.md
          done

      - name: Comment on PR
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: gh pr comment ${{ github.event.pull_request.number }} --body "$(cat combined_review.md)"

      - name: Add label
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: gh pr edit ${{ github.event.pull_request.number }} --add-label "ai-reviewed"

2. Configure Secrets

You must add the CODE_TURTLE_URL secret to your repository settings:

1. Go to Settings > Secrets and variables > Actions.
2. Click New repository secret.
3. Name: CODE_TURTLE_URL
4. Value: https://your-deployed-code-turtle-instance.com (e.g., your Railway or Azure app URL).

3. Usage

Once configured, Code Turtle will automatically run on:

• New Pull Requests.
• New commits pushed to an existing Pull Request.

It will analyze the diff, look for bugs and improvements, and post a consolidated review as a comment on the PR.

📄 License

MIT