{
"markdown": "# MCP Bio-Modelling Servers\n\n\n\n\n\n[](https://pypi.org/project/mcp-biomodelling-servers/)\n[](https://registry.modelcontextprotocol.io)\n\nThis package provides three stateful\n[Model Context Protocol](https://modelcontextprotocol.io/) servers for\nmechanistic and systems-biology modelling:\n\n| Server | Modelling role | Upstream project | MCP Registry name |\n|---|---|---|---|\n| MaBoSS | Configure, simulate, and analyze stochastic Boolean models | [pyMaBoSS](https://github.com/colomoto/pyMaBoSS) | `io.github.marcorusc/MaBoSS` |\n| NeKo | Build and analyze signalling networks from interaction databases | [NeKo](https://github.com/sysbio-curie/Neko) | `io.github.marcorusc/NeKo` |\n| PhysiCell | Build, inspect, and export PhysiCell and PhysiBoSS configuration files | [PhysiCell-settings](https://github.com/marcorusc/PhysiCell_Settings) | `io.github.marcorusc/PhysiCell` |\n\nAll three servers use MCP over stdio and are distributed together as\n`mcp-biomodelling-servers`.\n\n## Publication\n\nFor more details, please check the related article:\n\n> **\"Intelligent tool orchestration for rapid mechanistic model prototyping: MCP servers as AI-biology interfaces\"**
\n> Marco Ruscone, Miguel Vazquez & Alfonso Valencia, *npj Systems Biology and Applications* (2026)
\n> [https://doi.org/10.1038/s41540-026-00767-3](https://doi.org/10.1038/s41540-026-00767-3)\n\n## Requirements\n\n- Python 3.10–3.14.\n- MCP Python SDK 2.x, installed automatically with this package.\n- The modelling-package dependencies declared in `pyproject.toml`, installed\n automatically by `pip` or `uvx`.\n- The Graphviz system runtime for NeKo history diagrams. The Python `graphviz`\n package is not a replacement for the external `dot` renderer.\n\nCheck whether Graphviz is available with:\n\n```bash\ndot -V\n```\n\nIf this command is missing, install Graphviz using your operating system or\nenvironment package manager. See the\n[Graphviz installation guide](https://graphviz.org/download/) for\nplatform-specific instructions.\n\n## Installation\n\n### Install with pip\n\n```bash\npython -m pip install mcp-biomodelling-servers\n```\n\nThe installation provides three console entry points:\n\n```bash\nmcp-neko-server\nmcp-maboss-server\nmcp-physicell-server\n```\n\n### Run in an isolated environment with uvx\n\n```bash\nuvx --from mcp-biomodelling-servers mcp-neko-server\nuvx --from mcp-biomodelling-servers mcp-maboss-server\nuvx --from mcp-biomodelling-servers mcp-physicell-server\n```\n\nConda is optional. It remains useful when you want one explicitly managed\nenvironment for local development or additional native scientific software,\nbut it is not required for the packaged entry points.\n\n## Configure an MCP client\n\nThe following example uses `uvx` and works with clients that accept the common\n`mcp.json` stdio configuration:\n\n```jsonc\n{\n \"servers\": {\n \"neko\": {\n \"type\": \"stdio\",\n \"command\": \"uvx\",\n \"args\": [\n \"--from\",\n \"mcp-biomodelling-servers\",\n \"mcp-neko-server\"\n ]\n },\n \"maboss\": {\n \"type\": \"stdio\",\n \"command\": \"uvx\",\n \"args\": [\n \"--from\",\n \"mcp-biomodelling-servers\",\n \"mcp-maboss-server\"\n ]\n },\n \"physicell\": {\n \"type\": \"stdio\",\n \"command\": \"uvx\",\n \"args\": [\n \"--from\",\n \"mcp-biomodelling-servers\",\n \"mcp-physicell-server\"\n ]\n }\n }\n}\n```\n\nIf the package is already installed in the client environment, each entry can\ninstead use its console script directly:\n\n```jsonc\n{\n \"servers\": {\n \"neko\": {\n \"type\": \"stdio\",\n \"command\": \"mcp-neko-server\"\n },\n \"maboss\": {\n \"type\": \"stdio\",\n \"command\": \"mcp-maboss-server\"\n },\n \"physicell\": {\n \"type\": \"stdio\",\n \"command\": \"mcp-physicell-server\"\n }\n }\n}\n```\n\nRefer to your MCP client's documentation for its configuration-file location\nand reload procedure. For Visual Studio Code, see\n[Use MCP servers in VS Code](https://code.visualstudio.com/docs/copilot/chat/mcp-servers).\n\n## Sessions, artifacts, and errors\n\nEach server can maintain multiple isolated modelling sessions. Tools that\ncreate or load a model return a session identifier; pass that identifier to\nsubsequent operations when more than one session is active.\n\nGenerated models, configuration files, plots, and other outputs are kept in\nsession-scoped artifact directories. Artifact-listing tools return the paths\nneeded to inspect or hand files to another modelling server.\n\nUnder MCP SDK 2.x, failures to execute a tool are returned as tool errors so\nthe client and model can distinguish them from successful scientific results.\nValidation tools may still return a successful result describing an invalid\nmodel or configuration when validity itself is the requested result.\n\n## Run from source\n\nClone the repository and install it with its development dependencies:\n\n```bash\ngit clone https://github.com/marcorusc/mcp-biomodelling-servers.git\ncd mcp-biomodelling-servers\npython -m pip install \".[dev]\"\n```\n\nYou can then run the same console entry points or invoke a server module\ndirectly with the selected Python interpreter:\n\n```bash\npython MaBoSS/server.py\npython NeKo/server.py\npython PhysiCell/server.py\n```\n\n## Repository layout\n\n```text\nMaBoSS/ MaBoSS server, manual, and Registry manifest\nNeKo/ NeKo server, manual, and Registry manifest\nPhysiCell/ PhysiCell server, manual, and Registry manifest\nmcp_biomodelling_servers/ Installed package namespace and entry points\ntests/ Protocol, runtime, concurrency, and package tests\n```\n\nThe server-specific READMEs describe the modelling workflows and exposed tool\nfamilies in more detail.\n\n## MCP SDK and protocol compatibility\n\nThe package uses the stable MCP Python SDK 2.x API. The SDK negotiates the\nappropriate MCP protocol revision with the connected client; the protocol\nrevision is independent of the MCP Registry schema used by each `server.json`.\n\n## License\n\nThe package metadata declares the project under the MIT license. The wrapped\nmodelling packages retain their own licenses; consult their upstream projects\nfor details.\n",
"bytes": 6419,
"sha": "7158ec12663790bf27f659f663ec5e4e3f41aab10a8c1c0f6d4634cdc3d3b1b0",
"repo_slug": "marcorusc/mcp-biomodelling-servers",
"fonte": "repo",
"truncated": false,
"api": "https://agentalog.com/api/listings/mcp_io_github_marcorusc_maboss_44d3af53/readme"
}