tes-mcp-server

You run the tes-mcp-server MCP server to provide access to its API for AI agents and LLMs. It exposes weather and climate data through standardized MCP tools, with flexible deployment options including Docker and a local runtime, and supports asynchronous operations for responsive interactions.

How to use

Connect to the MCP server using an MCP client and start exploring the available tools. You can access the HTTP endpoint to interact with tes-mcp-server from any compatible client, or run a local stdio instance to embed the server directly into your workflow.

Common usage patterns include discovering available tools, invoking a tool with a query, and handling results asynchronously. Start by connecting to the server’s MCP URL, then enumerate tools, then call the tool you need for tasks such as obtaining weather data, forecasts, or climate information.

How to install

Prerequisites you need before installation are Python and a compatible runtime, Docker (for containerized deployment), and optionally Docker Compose for orchestration.

Step by step, you can install using one of the available deployment methods.

Option 1: Docker (Recommended) Run locally with Docker to quickly start the MCP server.

git clone https://github.com/Traia-IO/tes-mcp-server-mcp-server.git
cd tes-mcp-server-mcp-server
./run_local_docker.sh

Option 2: Docker Compose Create a .env file with configuration and start the services with Docker Compose.

PORT=8000
docker-compose up

Option 3: Manual Installation Install dependencies with uv and run the server directly.

uv pip install -e .
uv run python -m server

Additional setup and notes

Health checks and basic connectivity are important after starting the server. You can perform a health check to ensure the MCP server is running and reachable.

If you need to verify via a quick health test, run the health check script to confirm the server status.

You can connect to the HTTP endpoint at http://localhost:8000/mcp/ from an MCP client to begin using the tools exposed by tes-mcp-server.

Deployment and environment

Deployment supports Docker and Docker Compose for containerized setups. For cloud deployment, you can run the server under a compatible environment and expose the MCP endpoint to clients.

Environment variables you may configure include the port and stage, with a default port of 8000 and stage MAINNET. Adjust the configuration to fit your deployment needs.

Troubleshooting

If the server does not start, examine container or process logs for errors. For Docker-based runs, inspect container logs using your container management tooling. For local runs, check the console output for startup messages and exception details.

If you encounter connection issues, verify that the server is listening on the expected port and that the MCP URL matches your client configuration.

Tools and capabilities

The server exposes tools that allow interaction with tes-mcp-server endpoints through the MCP protocol. One placeholder example is example_tool, which is intended to be implemented and wired to the underlying API calls.

Available tools

example_tool

Placeholder tool (to be implemented)