Daytona MCP Interpreter

🚀 Daytona MCP Server Interpreter

Daytona MCP Server is a powerful tool for running and managing code in a controlled environment. It's particularly suitable for analyzing and executing code obtained from untrusted sources, such as code downloaded from the internet or scripts provided by users. The server offers a secure and isolated environment, ensuring that potentially malicious code won't damage the host system.

🚀 Quick Start

✨ Features

• Runs and manages code in a controlled environment.
• Analyzes and executes code from untrusted sources safely.
• Provides a secure and isolated environment to protect the host system.

📦 Installation

1. Install uv

To install the Daynta MCP Server, you first need to install uv:

npm install -g mcp-server

2. Create a Virtual Environment

It's recommended to run the Daynta MCP Server in a virtual environment:

Unix/MacOS:

python3 -m venv myenv
source myenv/bin/activate

Windows:

py -m venv myenv
myenv\Scripts\activate

3. Install Dependencies

Install the required dependencies:

pip install mcp-server

🔧 Technical Details

Environment Variables

The following environment variables may be used when running the Daynta MCP Server:

💻 Usage Examples

Development

1. Run the Server

Run the Daynta MCP Server:

mcp-server start

2. Test with MCP Inspector

Test using the MCP Inspector tool:

mcp-inspect http://localhost:3000

3. View Logs

View the server logs:

journalctl -u mcp-server.service -f

Integration

Configure the Daynta MCP interpreter in Claude Desktop:

1. Open Claude's settings interface.
2. Navigate to the "Interpreter" section.
3. Add a new interpreter, select "Custom path" and enter the following path:

http://localhost:3000

Available Tools

1. Shell Execution

Run shell commands in a controlled environment:

Basic Usage

shell_exec(command="echo 'Hello, World!'")

Advanced Usage

shell_exec(
    command="python3 -m http.server 8000",
    cwd="/workspace/my_project",
    capture_output=True,
    timeout=10
)

2. File Download

Download files from the controlled environment:

Basic Usage

file_download(remote_path="/workspace/example.txt")

Advanced Usage

file_download(
    remote_path="/workspace/data/config.json",
    local_path="config.json",
    overwrite=False,
    progress=True
)

3. File Upload

Upload files to the controlled environment:

Basic Usage

file_upload(local_path="example.txt", remote_path="/workspace/example.txt")

Advanced Usage

file_upload(
    local_path="local_image.png",
    remote_path="/workspace/images/uploaded.png",
    encoding="base64",
    overwrite=False
)

4. Git Clone

Clone a Git repository in the controlled environment:

Basic Usage

git_clone(repo_url="https://github.com/username/repository.git")

Advanced Usage

git_clone(
    repo_url="https://github.com/username/repository.git",
    branch="develop",
    target_path="my_project",
    depth=1,
    lfs=True
)

5. Web Preview

Generate a preview link for a web server running in the controlled environment:

Basic Usage

web_preview(port=8000, path="/")

Advanced Usage

web_preview(
    port=8000,
    path="/",
    title="My Web Application",
    description="Preview of my web application"
)

With these tools, you can securely run and manage code in a controlled environment, protecting the host system from potentially malicious code.