代码浏览器 - 框架代理

21-Day Agent Series: Day 11 AGENT : AI Code Review & Issue Fixer

AI Code Review & Issue Fixer This agent is part of the "A New AI Agent Every Day!" Series - Day 11/21 - AI Code Review & Issue Fixer 🛠🚀 This AI agent analyzes code repositories, detects potential security vulnerabilities, reviews code quality, and suggests fixes based on Sentry error logs using Sentry and GitHub MCP servers!

Installation

Prerequisites

• Python 3.9 or higher
• Git
• Virtual environment (recommended)

Steps

Don't forget to download nodejs for MCP

1. Clone the repository:

   git clone <repository-url>
   cd <repository-folder>
   

2. Install dependencies:

   pip install -r requirements.txt
   
   

3. Create a .env file in the root directory and configure it as follows:

   AZURE_OPENAI_ENDPOINT="your_azure_openai_endpoint"
   AZURE_OPENAI_API_VERSION="your_azure_openai_api_version"
   AZURE_OPENAI_API_KEY="your_azure_openai_api_key"
   GITHUB_PERSONAL_ACCESS_TOKEN="YOUR_GITHUB_TOKEN"
   SENTRY_AUTH_TOKEN="YOUR_SENTRY_TOKEN"
   

Running the Application

Start the FastAPI server:

uvicorn upsonicai:app --reload

Open the UI in your browser:

http://127.0.0.1:8000/

MCP Configuration

Modify your MCP configuration to include GitHub and Sentry:

"mcpServers": {
  "github": {
    "command": "npx",
    "args": ["-y", "@modelcontextprotocol/server-github"],
    "env": {
      "GITHUB_PERSONAL_ACCESS_TOKEN": "YOUR_GITHUB_TOKEN"
    }
  },
  "sentry": {
    "command": "python",
    "args": ["-m", "mcp_server_sentry", "--auth-token", "YOUR_SENTRY_TOKEN"]
  }
}

How It Works

• Fetches recent commits from GitHub/GitLab repositories
• Analyzes code quality and detects security vulnerabilities
• Retrieves error logs from Sentry
• Provides actionable insights and fixes for detected issues

API Documentation

Interactive API docs are available at:

• Swagger UI: http://127.0.0.1:8000/docs
• ReDoc: http://127.0.0.1:8000/redoc