mcp-serve

# mcp-serve > Agentic Server to support MCP Tools and Science  [](https://doi.org/10.5281/zenodo.19476301) [](https://badge.fury.io/py/mcp-serve) ## Design This is a customizable, asynchronous server that can register and load tools of interest. Endpoints include functions (tools), prompts, resources. Those are now (for the most part) separated into modular projects: - [flux-mcp](https://github.com/converged-computing/flux-mcp): MCP tools for Flux Framework - [hpc-mcp](https://github.com/converged-computing/hpc-mcp): HPC tools for a larger set of HPC and converged computing use cases. ### Abstractions The library here has the following abstractions. - **tools**: server tools, prompts, and resources - **ui**: user interface that an engine (with a main manager) uses - **core**: shared assets, primarily the plan/step/config definitions and worker/hub hierarchy roles - **routes**: server views not related to mcp. - **databases**: how to save results as we progress in a pipeline (currently we support sqlite and filesystem JSON) For the above, the engines, tools, ui, databases, and backends are interfaces. ## Development It is recommended to open in VSCode container. Then install: ```bash pip install --break-system-packages -e . ``` ### Docker To build the Docker container: ```bash make ``` To run a dummy example: ```bash docker run -p 8089:8089 -it ghcr.io/converged-computing/mcp-server:latest ``` And then interact from the outside: ```bash python3 ./examples/echo/test_echo.py ``` ### Environment The following variables can be set in the environment. | Name | Description | Default | |-------|------------|---------------| | `MCPSERVER_PORT` | Port to run MCP server on, i…