Devin.cursorrules: Enhancing the Cursor IDE Development Experience, Making Cursor/Windsurf a Devin-like Intelligent Programming Assistant

General Introduction

Devin.cursorrules is an innovative open source project designed to convert the affordable Cursor or Windsurf IDEs (~$20) into powerful AI programming assistants like Devin. The project provides a complete set of profiles and toolsets that can infuse the IDE with intelligent agent AI functionality to significantly improve development efficiency. Core features include intelligent process planning, self-evolving capabilities, extended tool usage (support for web browsing, search, and LLM-based analytics), and automated execution in Docker containers (for Windsurf users). The project is licensed under the MIT license, which allows developers to use and modify it freely. With simple configuration, developers can upgrade the basic IDE to a powerful development tool with advanced AI assistant features in just a minute.

Related presentation:Transform Cursor and Windsurf into $500/month worth of Devin in an hour!

SVG enlargement process

Function List

• Intelligent process planning and self-evolution capabilities
• Web crawler (based on Playwright, supports JavaScript rendering)
• Search engine integration (integrated DuckDuckGo search)
• LLM-driven text analysis (with support for custom models)
• Automated execution environment (Windsurf Docker container support)
• Complete unit test suite
• Cross-platform support (Windows/Unix/macOS)
• Modular Tool Integration System
• Advanced Web Content Extraction and Processing
• Asynchronous Concurrent Request Processing
• DEBUG logging system

Using Help

1. Environmental configuration

First you need to configure the Python virtual environment to ensure that project dependencies do not affect the system environment:

# Creating a Python virtual environment
python3 -m venv py310
# Activating the virtual environment
# Windows systems.
. \py310\Scripts\activate
# Unix/macOS systems: .
source py310/bin/activate

2. Installation of necessary components

The project depends on the installation process:

# installs the required packages
pip install -r requirements.txt
# Install Playwright's Chromium browser (for web crawling)
python -m playwright install chromium

3. Configuration file deployment

1. Copy all files from the repository to your project directory
2. Cursor users:
   • .cursorrulesThe file is automatically loaded
   • No additional configuration required
3. Windsurf users:
   • Requires concurrent use.windsurfrulescap (a poem)scratchpad.md
   • Make sure the file is in the correct directory location

4. Description of the use of functional modules

4.1 Web Page Capture Function

• Supports JavaScript-rendered web content crawling
• Automatic handling of coding issues and support for internationalized content
• Concurrent request processing for efficiency
• Example of use:

from tools.web_scraper import process_urls
urls = ["https://example.com"]
results = await process_urls(urls)

4.2 Search Engine Integration

• Using the DuckDuckGo backend
• Support for customizing the number of search results
• Provide detailed information on search results
• Example of use:

from tools.search_engine import search
search("your query", max_results=10)

4.3 LLM Integration

• Support for customized LLM servers
• Default use of Qwen 2.5-32B model
• Configurable temperature parameters
• Example of use:

from tools.llm_api import query_llm
response = query_llm("your prompt")

5. Testing and validation

The program offers a complete test suite:

# Ensure that you are in a virtual environment
source py310/bin/activate
# Run all tests
PYTHONPATH=. python -m unittest discover tests/

6. Troubleshooting

• If the LLM service is not configured, the relevant features are gracefully degraded
• Network problems can be seen in the DEBUG message of stderr
• Ensure Python version compatibility (Python 3.10+ recommended)
• Check that the dependency package versions are correct