opencursor 0.0.21

An AI-powered code agent for workspace operations

OpenCursor

An AI-powered code agent for workspace operations with a rich terminal UI.

Overview

OpenCursor is a terminal-based AI coding assistant that helps you navigate, understand, and modify codebases. It provides both autonomous and interactive agent modes, along with direct LLM chat capabilities. The tool uses a variety of AI-powered features to help you work with code more efficiently.

Features

🤖 Local LLM Support - Works with Ollama models locally

🔄 Autonomous Mode - Complete tasks without intervention

🌐 Playwright Web Search - No API limits or blocking

📝 Full Editor Toolkit - Search, edit, delete files & run commands

🔍 Transparent UI - See exactly what the model is doing

Example Queries

• "Create a simple Flask API with user authentication"
• "Refactor this React component to use hooks instead of classes"
• "Find all usages of this function and update its parameters"
• "Analyze this codebase and explain its architecture"
• "Debug why this test is failing and propose a fix"
• "Research and implement the latest best practices for API security"

Installation

Using pip (recommended)

pip install -U opencursor

Using Poetry

# Clone the repository
git clone https://github.com/santhoshkammari/OpenCursor.git
cd OpenCursor

# Install with Poetry
poetry install

Usage

Once installed, you can use OpenCursor from the command line:

# Basic usage
opencursor

# Specify a workspace directory
opencursor -w /path/to/workspace

# Use a different model
opencursor -m "gpt-4"

# Start with an initial query
opencursor -q "Create a simple Flask app"

Command-line Options

• -w, --workspace: Path to the workspace directory (default: current directory)
• -q, --query: Initial query to process
• -m, --model: LLM model to use (default: qwen3_14b_q6k:latest)
• -H, --host: Ollama API host URL (default: http://192.168.170.76:11434)
• --no-thinking: Disable thinking process in responses

Commands

OpenCursor provides several commands that you can use within the application:

• /agent <message>: Send a message to the agent (autonomous mode)
• /interactive <message>: Send a message to the agent (interactive mode)
• /chat <message>: Chat with the LLM directly (no tools)
• /add <filepath>: Add a file to the chat context
• /drop <filepath>: Remove a file from the chat context
• /clear: Clear all files from the chat context
• /repomap: Show a map of the repository
• /focus <filepath>: Focus on a specific file
• /diff <filepath>: Show git diff for a file with syntax highlighting
• /help: Show help information
• /exit: Exit the application

You can also use shortcuts:

• @filepath to quickly add a file to the context

Development

Project Structure

opencursor/
├── code_agent/
│   ├── src/
│   │   ├── app.py         # Main application with UI
│   │   ├── agent.py       # Agent implementation
│   │   ├── llm.py         # LLM client
│   │   ├── tools.py       # Tool implementations
│   │   ├── prompts.py     # System prompts
│   │   ├── tool_playwright.py  # Web search tools
│   │   └── tool_browser.py     # Browser tools
│   ├── cli_entry.py       # CLI entry point
│   └── __init__.py
├── pyproject.toml         # Poetry configuration
├── requirements.txt       # Dependencies
└── README.md

Core Components

1. OpenCursorApp: Main application class that handles the UI and command processing
2. CodeAgent: Handles autonomous and interactive modes, manages tool execution
3. LLMClient: Interacts with the Ollama API, manages conversation history
4. Tools: Implements various tools for file operations, code analysis, etc.

Setting Up Development Environment

# Clone the repository
git clone https://github.com/santhoshkammari/OpenCursor.git
cd OpenCursor

# Install dependencies
pip install -e .
# or with poetry
poetry install

# Run the application
python -m code_agent.cli_entry

Dependencies

• Python 3.11+
• Rich: Terminal UI and formatting
• Ollama: LLM API client
• Prompt_toolkit: Command completion and input handling
• Playwright: Web search functionality
• SentenceTransformer: Semantic code search

License

MIT

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add some amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request