An MCP server to upload your projects to yourware.so
Yourware MCP is a Model Context Protocol (MCP) server that enables AI assistants like Claude to upload your projects directly to yourware.so. This integration allows for seamless deployment of web projects, making it easy to showcase your work online without leaving your development environment.
Whether you're building a simple HTML page, a complex web application, or any other web-based project, Yourware MCP streamlines the process of getting your work online and shareable.
Simple API key-based authentication system that securely stores your credentials.
Upload single files or entire directories with automatic compression and organization.
One-step deployment process that makes your projects instantly available online.
Works with AI assistants like Claude through Cline, Cursor, and Windsurf for a smooth workflow.
Getting started with Yourware MCP is straightforward:
pip install yourware-mcp
Or use uvx
directly as shown in the configuration examples below.
Add the Yourware MCP server to your AI assistant's configuration:
{ "mcpServers": { "yourware-mcp": { "command": "uvx", "args": ["yourware-mcp@latest", "stdio"], "env": {} } } }
In Cursor settings → features → MCP Servers, add a new MCP Server named yourware-mcp
and set the command to uvx yourware-mcp@latest stdio
You can either:
YOURWARE_API_KEY
environment variableOnce configured, you can use Yourware MCP through your AI assistant with the following tools:
// Check if your credentials exist and are valid await use_mcp_tool({ server_name: "yourware-mcp", tool_name: "check_credentials", arguments: {} });
// Create a new API key await use_mcp_tool({ server_name: "yourware-mcp", tool_name: "create_api_key", arguments: { api_key: "your-api-key-here" // Optional, if not provided, will guide you to create one } });
// Upload a file or directory to yourware await use_mcp_tool({ server_name: "yourware-mcp", tool_name: "upload_project", arguments: { file_path: "path/to/your/project" // Can be a directory or single file } });
A typical workflow might look like this: