BioPB

Performing image analysis by visiting biopb endpoints

Ji Yu

biopb-mcp is the MCP component of the BioPB project

Installation - recommended

curl -fsSL https://biopb.org/install.sh | bash

Installation - alternative methods

These methods only install the mcp component without the rest of BioPB system.

From PyPI

pip install biopb-mcp        # or: uv pip install biopb-mcp

Standalone PyInstaller app bundles (the per-platform .zip/.tar.gz downloads) are temporarily disabled — installation is uv/pip-based for now. Use the recommended one-liner above, or the methods here.

Development Version

To install latest development version:

pip install &quot;git+https://github.com/biopb/biopb.git#subdirectory=biopb-mcp&quot;

Troubleshooting

For URL resolution, the auto-start fallback, and how to resolve startup failures (e.g. a port already in use on a shared node), see docs/troubleshooting.md.

Contributing

Contributions are very welcome. Read development.md first to understand the project architecture.

License

Distributed under the terms of the MIT license, "biopb-mcp" is free and open source software

Issues

If you encounter any problems, please file an issue along with a detailed description.

Version:

0.7.2

Last updated:

2026-06-13

First released:

2026-05-29

License:

Supported data:

Information not submitted

Plugin type:

Widget

Open extension:

Save extension:

Python versions supported:

Operating system:

Information not submitted

Requirements:

numpy
pandas
magicgui
qtpy
scikit-image
biopb[tensor]>=0.6.4
opencv-python-headless
grpcio-health-checking
napari-skimage-regionprops; extra == "tables"
mcp>=1.20; extra == "mcp"
uvicorn>=0.29; extra == "mcp"
jupyter_client; extra == "mcp"
ipykernel; extra == "mcp"
psutil; extra == "mcp"
distributed; extra == "mcp"
napari>=0.7; extra == "mcp"
pyqt6; extra == "mcp"
vedo; extra == "3d"