alara 0.3.3

A stdio bridge for the Alara MCP API

Alara Bridge (StdIO)

This package provides a Machine Cognition Protocol (MCP) bridge using standard input/output (stdio) to connect an MCP client (like Cursor) to the Alara backend API.

Features

• Dynamically fetches available tools from the Alara OpenAPI schema.
• Executes tool calls by forwarding requests to the Alara API.
• Authenticates with the backend using an API key provided via environment variable.

Installation

# Install from PyPI
pip install alara

Configuration

This bridge uses the command-based configuration method for MCP clients like Cursor. The package includes a helper script to generate the necessary JSON snippet for you.

1. Install the package:

   pip install alara
   

2. Generate your API Key: Obtain your API key from the Alara platform dashboard.

3. Generate MCP Configuration: Run the following command in your terminal, replacing YOUR_ALARA_API_KEY_HERE with the key you obtained:

   python -m alara.main --api-key YOUR_ALARA_API_KEY_HERE --print-mcp-config
   

   (Note: If python doesn't work, try python3)

4. Copy the Output: The command will print a JSON object similar to the example below. Copy this entire object.

   {
       "mcpServers": {
           "alara": {
               "name": "Alara",
               "description": "Alara Bridge (Python Module)",
               "protocol": "stdio",
               "command": "/path/to/installed/python", // Path to python in venv
               "args": ["-m", "alara.main"],
               "cwd": "/path/to/alara/project/root",
               "env": {
                   "ALARA_API_KEY": "YOUR_ALARA_API_KEY_HERE", // Your key inserted here
                   "ALARA_MCP_URL": "https://alara-mcp.skolp.com" // Changed default backend URL
               }
           }
       }
   }
   

   (The command and cwd paths will be automatically detected based on your installation.)

5. Configure Cursor: Open your Cursor configuration file (e.g., ~/.cursor/mcp.json) and paste the copied JSON object into the "mcpServers" section (or merge it if the section already exists).

6. Reload MCP Clients: Reload the clients in Cursor (e.g., via the command palette).

Development Setup

1. Clone the repository:

   git clone https://github.com/rizkisyaf/alara.git
   cd alara
   

2. Create a Python virtual environment:

   python3 -m venv venv
   source venv/bin/activate 
   

3. Install dependencies, including development tools:

   pip install -e ".[dev]" 
   

   (Note: Requires defining [project.optional-dependencies] in pyproject.toml if you have dev tools like pytest, ruff, etc.)

4. Create a .env file in the project root with your API key for local testing:

   ALARA_API_KEY=YOUR_ALARA_API_KEY_HERE
   ALARA_MCP_URL=https://alara-mcp.skolp.com # Changed default backend URL
   # Optional: Set LOG_LEVEL=DEBUG for more verbose logging
   # LOG_LEVEL=DEBUG 
   

5. Run the bridge directly (for testing purposes):

   python -m alara.main 
   

License

This project is licensed under the MIT License - see the LICENSE file for details (if one exists).