d0ng231

Open Med MCP — Development skill for Claude Code

Development community

Open, modular MCP server for medical image analysis agents: containerized models (MedSAM2, SAM 2.1, TotalSegmentator, lungmask, HD-BET, SynthStrip, nnU-Net, MONAI, TorchXRayVision), preset/custom guid.

How to install Open Med MCP

This entry records only its repository, not the path inside it, so there is no exact command to give. Open d0ng231/open-med-mcp and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Open Med MCP does

Alternatives in Development

Prompt Net Requests — I want to enhance the AsyncPlaywrightCrawlerStrategy to optionally capture network requests and console messag 85.8k ★
Package OfficeCLI Skill As AionUi Assistant — Convert an OfficeCLI skill into a fully wired AionUi assistant preset, or update an existing one 20.3k ★
Market Sizing — Estimate TAM, SAM, SOM with top-down and bottom-up approaches 7.8k ★

README

**open-med-mcp** is an open, modular [Model Context Protocol](https://modelcontextprotocol.io) server that lets AI coding agents (Claude Code, Codex CLI, Claude Desktop, Cursor, ...) do **end-to-end medical image analysis** on your machine: inspect a CT/MRI/X-ray, pick a protocol, run a segmentation model, *look* at the result, refine it, measure it, and write a report.

It is built on three pillars:

pillar what it means

1. Containerized specialised models Every model is an adapter with a manifest, a run.py and a Dockerfile, speaking one tiny job-directory contract. Run it locally (pip extra), in Docker (GHCR images) or in Apptainer on an HPC cluster - the agent does not care which. Ships with MedSAM2, SAM 2.1, TotalSegmentator, lungmask, HD-BET, SynthStrip (official FreeSurfer image), nnU-Net (any trained model), MONAI Model Zoo bundles, TorchXRayVision and a classical baseline; adding your own model is a folder with three files, and existing third-party images can be wrapped with a command template.

2. Preset / custom guidelines Markdown protocols with YAML front matter that tell an agent how to do a task well (which model, which window, what to check, what to report). Presets ship with the packa

pillar	what it means
1. Containerized specialised models	Every model is an adapter with a manifest, a `run.py` and a Dockerfile, speaking one tiny job-directory contract. Run it locally (pip extra), in Docker (GHCR images) or in Apptainer on an HPC cluster - the agent does not care which. Ships with MedSAM2, SAM 2.1, TotalSegmentator, lungmask, HD-BET, SynthStrip (official FreeSurfer image), nnU-Net (any trained model), MONAI Model Zoo bundles, TorchXRayVision and a classical baseline; adding your own model is a folder with three files, and existing third-party images can be wrapped with a command template.
2. Preset / custom guidelines	Markdown protocols with YAML front matter that tell an agent how to do a task well (which model, which window, what to check, what to report). Presets ship with the packa