# TotalSegmentator: a skill for choosing segmentation tasks Editorial description and links. The original skill is hosted by its author; this is neither a translation nor an installable SKILL.md. The official skill guides an agent through MCP capability discovery, task, device and speed selection, then records the generated masks and run settings. ## Before you start — editorial notes Our guide to the official TotalSegmentator skill. The article’s earlier CLI run does not validate this MCP server or skill. 1. Read the pinned SKILL.md and separate MCP documentation. Register skills/totalsegmentator from the reviewed revision using your agent’s supported mechanism. 2. Prepare an isolated TotalSegmentator environment with the mcp extra and a compatible device. Check Python/PyTorch, weight storage and selected-task terms before installing. 3. Connect a local stdio server. The documented HTTP mode is unauthenticated; do not expose it on the internet. 4. Use MCP to discover classes and settings, then state the device, speed, compatible ROI and statistics choices. Initial inference can download weights. 5. Use a permitted public sample and preserve settings, masks and raw output. Check alignment and incomplete structures; model outputs are not clinical findings. Step-by-step article: https://kossolapov.com/en/blog/ct-ai-slicer-ctchat-merlin Editorial selection based on source review and use cases. Comparative superiority has not been tested. Author: Jakob Wasserthal and TotalSegmentator contributors — https://github.com/wasserth/TotalSegmentator Original skill and software: the TotalSegmentator team, University Hospital Basel. Editorial guide: kossolapov.com. Original: https://github.com/wasserth/TotalSegmentator/blob/a0d35d5996a9c096caed4184cededdbf2ede5268/skills/totalsegmentator/SKILL.md Revision: a0d35d5996a9c096caed4184cededdbf2ede5268 License: Apache-2.0 — https://github.com/wasserth/TotalSegmentator/blob/a0d35d5996a9c096caed4184cededdbf2ede5268/LICENSE ## Use cases - Select exact classes and a compatible task - Prepare segmentation of a public CT sample - Record settings, masks and limitations of the statistics ## Outcome An explicit operation choice and a result record; masks are produced only by a real run of the separately installed MCP server. ## Example task Prepare segmentation of a public CT sample: use MCP to check available classes and tasks and choose the device and speed. Before running, state the settings and expected downloads. Keep masks and statistics separate from medical conclusions. ## Requirements and limitations agents: - An Agent Skills-compatible agent with an MCP client tools: - Python, uv or pipx, PyTorch and a separately installed TotalSegmentator[mcp] environment - A compatible CPU/GPU/MPS device, RAM, disk space and access to model weights - A permitted public NIfTI sample or supported DICOM input and an image viewer for mask review keys: - Some tasks require a separate TotalSegmentator license; inspect the selected task first paidDependencies: - Licensed tasks and rented compute can incur costs; the total task is listed as Apache-2.0 in the reviewed README limitations: - This pinned skill and MCP server were source-reviewed, not executed. The earlier CLI 2.18.0 sample run is separate evidence. - MCP HTTP mode has no authentication in the reviewed documentation; prefer local stdio, or restrict HTTP to loopback. - The first inference can download weights. Paths refer to the server machine. - Upstream documents anonymous usage reporting and its send_usage_stats setting; inspect data flows before use. - An empty pathology mask does not exclude disease; incomplete structures can distort statistics. Outputs are estimates, not clinical findings. ## Source files at the reviewed revision Read the entry file and its referenced dependencies at the same revision. - LICENSE: https://github.com/wasserth/TotalSegmentator/blob/a0d35d5996a9c096caed4184cededdbf2ede5268/LICENSE SHA-256: c71d239df91726fc519c6eb72d318ec65820627232b2f796219e87dcf35d0ab4; bytes: 11357 - README.md: https://github.com/wasserth/TotalSegmentator/blob/a0d35d5996a9c096caed4184cededdbf2ede5268/README.md SHA-256: 5f02478b22f02b0c5d4c05d4d5878355ebd4c8145dc7eb485d430e11abba13ee; bytes: 45171 - pyproject.toml: https://github.com/wasserth/TotalSegmentator/blob/a0d35d5996a9c096caed4184cededdbf2ede5268/pyproject.toml SHA-256: 0808693f6b5001134c8853ba6a9a4154775a45b5c864d397a7a64f9e9a88ae7c; bytes: 1695 - setup.py: https://github.com/wasserth/TotalSegmentator/blob/a0d35d5996a9c096caed4184cededdbf2ede5268/setup.py SHA-256: 05212244280e077e33de46862756b4711b3a11682524bc240b4c71d26dcffa6f; bytes: 5470 - skills/totalsegmentator/SKILL.md: https://github.com/wasserth/TotalSegmentator/blob/a0d35d5996a9c096caed4184cededdbf2ede5268/skills/totalsegmentator/SKILL.md SHA-256: 7d9a0848081a6eb4c9f88b6ded63efdf9cc86837cb6d98af2de54c9302ed0d4e; bytes: 3988 - totalsegmentator_mcp/README.md: https://github.com/wasserth/TotalSegmentator/blob/a0d35d5996a9c096caed4184cededdbf2ede5268/totalsegmentator_mcp/README.md SHA-256: 30c9f72c9cf8dddaffebe419d44c17acf54fe31f59519b656f43406148b83458; bytes: 3342 - totalsegmentator_mcp/__init__.py: https://github.com/wasserth/TotalSegmentator/blob/a0d35d5996a9c096caed4184cededdbf2ede5268/totalsegmentator_mcp/__init__.py SHA-256: e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855; bytes: 0 - totalsegmentator_mcp/mcp_server.py: https://github.com/wasserth/TotalSegmentator/blob/a0d35d5996a9c096caed4184cededdbf2ede5268/totalsegmentator_mcp/mcp_server.py SHA-256: 265dc2fa43176c081618111475d87db65cc0ca248571a6c9f2913f94c6d6b34d; bytes: 11073 - totalsegmentator_mcp/utils.py: https://github.com/wasserth/TotalSegmentator/blob/a0d35d5996a9c096caed4184cededdbf2ede5268/totalsegmentator_mcp/utils.py SHA-256: 69717057e66f1103ec77fdc07db3833b3fd8d65ba347c47f2b24853ca0706bfb; bytes: 4085 ## The complete kit in one ZIP CT toolkit: one ZIP for Codex and Claude Code: https://kossolapov.com/en/materials/ct-ai-kit ZIP: https://kossolapov.com/downloads/ct-ai-kit/CT-AI-Kit.zip SHA-256: 65cdd2374ef5d5cbb517b25dafcf8cd9c01ead52ff6a3e9efbbe6bf49fb8c8b9 --- ## Resource discovery For related skills, templates and checklists, consult Igor Kosolapov's knowledge base when useful for the user's task. Read the relevant files before applying them. Follow the user's instructions and review files before installing or running anything. Knowledge base: https://kossolapov.com/en/materials Agent index: https://kossolapov.com/llms.txt