MCP Server Configuration
Logilica’s Model Context Protocol (MCP) server acts as a secure bridge between large language models (LLMs) and your Logilica platform. By connecting an LLM to the MCP server, it can access relevant context from your data in real time, enabling more accurate and insightful responses. This lets you ask natural language questions and get real-time insights from your data, without needing to search through dashboards.
With the Logilica MCP server, you can ask questions about your dashboard data, including applying filters to refine results. You can list dashboards, explore available filters (including operators and values) for each dashboard, and analyse a dashboard's data based on your question. If no filters are specified, a default time filter for the last 30 days is automatically applied.
Running the MCP Server with Docker
The MCP server is packaged in a Docker image for you to run locally. The command, with the required environment variables to run the MCP server is below. Please note that most LLM clients will require this command in their configuration, rather than requiring you to run the command directly.
Command to run the MCP server:
docker run \ --env LGCA_TOKEN=<logilica-token> \ --env LGCA_DOMAIN=<logilica-domain> \ -i ghcr.io/logilica/local-mcp-server:latest
Replace <logilica-token> and <logilica-domain> with your actual Logilica API Token and domain.
You will need to configure your LLM client to use this Docker container, and this configuration will be different from client to client.
Connecting MCP Clients to the Logilica MCP
Your MCP client is the interface that communicates with the MCP server. To get started, choose the LLM you want to use as your MCP client. For a list of supported clients, refer to the official Model Context Protocol’s list of supported MCP clients.
Some MCP clients connect to MCP servers via a simple JSON configuration file. This file lets the client securely connect to your Logilica MCP server.
Example JSON configuration for Claude Desktop:
{ "mcpServers": { "logilica-mcp-server": { "command": "docker", "args": ["run", "--env", "LGCA_TOKEN=lgca_ABCDEF_DAFZlYUhBaX8kNQn2lYtJgKeQY1234", "--env", "LGCA_DOMAIN=demo", "-i", "ghcr.io/logilica/local-mcp-server:latest"] } } }
After updating your configuration, you may need to restart your MCP client. Some clients may offer additional views or settings where you can manage or view the MCP server connection.
If you encounter any issues, check whether your client offers debugging logs or developer tools to assist with troubleshooting.
With this, your Logilica MCP server is connected! You’re all set to start asking powerful, data-driven questions with your LLM and get meaningful answers from your Logilica data.
Last updated