galaxy-mcp: the Galaxy bioinformatics platform in an agent
galaxy-mcp
Galaxy project MCP server connects an agent to a Galaxy instance: tool search and runs, histories, workflows and the IWC catalog
- 40 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The Galaxy API key gives access to the user histories, and the agent can run jobs, change and delete histories and upload files
- Running tools and workflows uses the Galaxy server compute and the user quota
- Research data, clinical data included, enters the model context when history contents are read
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add mcp/galaxy-mcpDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The Galaxy API key gives access to the user histories, and the agent can run jobs, change and delete histories and upload files
- Running tools and workflows uses the Galaxy server compute and the user quota
- Research data, clinical data included, enters the model context when history contents are read
This is third-party code. Review the repository files before installing.
What it does
The server connects to any Galaxy instance by URL and API key and gives the agent tools for the bioinformatics platform. It can search tools by name and keyword, inspect their inputs, run them in a history and follow jobs with stdout and stderr logs. Separate tool groups manage histories, datasets and collections, upload files from disk or by URL and download results. For workflows the agent lists them, builds an input template, invokes, lists invocations and cancels them. The Intergalactic Workflow Commission catalog adds search, ranking by task description and import of ready workflows. There are Galaxy pages in markdown and user-defined tools that are created and run without admin rights. The server ships in two implementations: Python with the galaxy-mcp package on PyPI and TypeScript with npm packages.
Who it is for. Bioinformaticians and researchers on Galaxy who want to run analyses, find workflows and investigate failed jobs by talking to an agent
Good fit when
- For sequencing data you search the IWC catalog for a suitable workflow and import it into your Galaxy
- A job in a history failed and the cause is investigated through its exit code and stdout and stderr logs
- Files from disk or by URL are uploaded to a history and a tool runs on them without the web UI
- An input template is built for a tool and its tests are checked for how parameters are set
- A workflow gets its inputs checked, is invoked and its progress is followed
Not a fit when
- You do not use Galaxy: the server talks only to servers of this platform
- You have no API key or OAuth sign-in: without them the connection to a Galaxy server is impossible
- You need the whole tool catalog in one request: the tool panel on a production server is megabytes in size
- You need reports of results for publication: the server runs analyses and downloads data, and formatting the report is up to you
Example request
Connect to usegalaxy.org, find an RNA-seq workflow in the IWC catalog, import it and show which inputs are needed to run it.MCP
- Transport
- stdio
- Authentication
- not required
Environment variables
GALAXY_URLrequired- The Galaxy server URL, for example https://usegalaxy.org/
GALAXY_API_KEYsecret- The Galaxy API key for stdio mode, not needed with OAuth sign-in
Limitations
The Python and TypeScript implementations live separately with different release trains, so their operation sets differ, and you should check the README of the one you installed. Some operations, such as page history, need Galaxy 26.1 or newer. The container-recommend extra is installed separately and adds dependencies. OAuth mode over HTTP requires a persistent session secret. The last commit was made on November 5, 2026.
How to disable. Run claude mcp remove galaxy and delete the API key on the Galaxy user page. Data in histories stays on the server and is removed with Galaxy itself.
FAQ
Do I need my own Galaxy server?
No. The connection goes to any reachable server, for example usegalaxy.org or regional ones, and needs its URL and API key.
How do I start the server?
For Python with uvx galaxy-mcp in stdio mode, and for HTTP add the transport parameter with the value streamable-http. A container image is on ghcr.io.
What does the IWC catalog give?
Search of ready Intergalactic Workflow Commission workflows, ranking by task description and import of the chosen workflow into your Galaxy.
Can I sign in without an API key?
Yes, in HTTP mode with OAuth the server exchanges credentials for temporary Galaxy keys.
What is the container-recommend extra for?
It resolves a real quay.io/biocontainers image for a set of conda packages so a user-defined tool does not fail on an invented tag.
Discussion
No questions or reviews yet.
Related
MCP serversAn MCP server with a toolbox for global briefs: markets, SEC filings, conflict, climate and news from open sources
A Claude Code plugin and MCP server: over 1000 scientific tools for biology, chemistry and medicine, data-check commands and routing skills
The official Graphlit MCP server: an agent collects material from Slack, GitHub, mail and websites into a knowledge base and searches it
The MCP server, CLI and SDKs of the QVeris platform: the agent finds a data service, calls it and audits credit spending
An agent reads Weights & Biases runs, metrics and Weave traces and builds reports from plain-language questions
An MCP server of science agents: Claude Code analyzes spectra and images, plans experiments and prepares DFT calculations
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| galaxy-mcp: the Galaxy bioinformatics platform in an agent (this card) | yes | not required | 40 | |
| BioMCP | no | not required | 678 | access to biomedical databases and literature, while this runs analyses and workflows on a Galaxy server |
| ClawBio | no | 1,182 | a set of bioinformatics agent skills, while this connects to the Galaxy platform |
README in short
The README describes the two implementations and the contract between them, the feature list from connection and tool search to pages and the IWC catalog, the container-recommend and code-mode extras, the stdio and HTTP modes with OAuth, and running in a container. The detailed operation list sits in the README of each implementation.