M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
G

Galaxy MCP Server

Official
MCP server for the Galaxy bioinformatics platform
How to install ↓ Compare GitHub source repository ↗
Category
Dev Tools #108 of 438
Stars
★ 39 Popular
Transport
stdio (local process) · Streamable HTTP
Runtime
Python · Docker · Node.js
Credentials
API key / credential required
License
MIT
Last commit
61FMRS · C

An officially maintained, MIT-licensed MCP server from the Galaxy Project built on BioBlend. It provides Galaxy connection, tool search/execution, workflow and history management, file upload, stdio and streamable-http transports, optional OAuth login, and a full test suite. Best suited for Galaxy users bringing the platform into AI assistants, provided they have Galaxy access credentials.

Strongest · Documentation 14/20 Weakest · Reliability 9/20

Reliability
9/20
Security and permissions
13/20
Maintenance
12/20
Documentation
14/20
Setup experience
13/20
Why each score
Reliability 9/20
The manifest declares only a single pypi package (stdio transport via uvx, clear command and env vars), but the README describes a substantially broader surface: HTTP/streamable-http transport, an OAuth flow exchanging credentials for short-lived Galaxy API keys, per-session credential isolation, and a parallel TypeScript implementation (galaxy-agent-tools) kept 'in lockstep' — none of which is reflected in server.json's packages, creating a manifest/README capability mismatch. The README claims a 'full test suite with mock-based testing' but no test files or CI workflow are provided as verifiable evidence. Per the static calibration rule, reliability is capped at 12 absent verifiable in-repo CI/test evidence; the unverified test claim plus the manifest/README inconsistency pushes the score down further to 9.
Security and permissions 13/20
The API key env var is correctly flagged isSecret, the OAuth flow trades long-lived credentials for short-lived Galaxy API keys, and the README describes per-session credential isolation for multi-user HTTP deployments — all positive security signals. Docker/config examples use placeholder values rather than real secrets, and no red-line issues were found (no malware, no covert exfiltration, no fabricated identity or test results). However, the server can execute Galaxy tools, manage workflows, upload files, and connect to any user-specified Galaxy instance, and the README does not describe confirmation gating for potentially destructive operations (e.g., deleting histories, running workflows) nor fully disclose data flow between local files, credentials, and the external Galaxy server, which caps the score below full marks.
Maintenance 12/20
The repository belongs to the established galaxyproject bioinformatics organization, uses the MIT license, and shows an iterative version (1.4.0); it is not archived, and has 37 stars with 19 open issues indicating some activity. However, the provided material has no changelog, no explicit release cadence, no documented security-response channel or maintainer identity, and no CI configuration to verify continuous integration, so it lands as 'active but with governance gaps' rather than higher.
Documentation 14/20
The README is reasonably well structured, covering an overview, quick start for both stdio and HTTP+OAuth modes, PyPI install, source build, Docker usage, and a concrete Claude Desktop configuration example, plus pointers to deeper Python and TypeScript sub-project docs. However, the material provided lacks tool/parameter-level documentation (tool list, input/output schemas), rate-limit or cost disclosures, and a dedicated troubleshooting section, placing it at a 'usable but with documentation gaps' level.
Setup experience 13/20
The basic stdio setup path is clear: a single uvx command works, with pip install, source build, Docker, and a concrete Claude Desktop config example all provided for mainstream client integration. The HTTP/OAuth mode, however, introduces additional environment variables (PUBLIC_URL, SESSION_SECRET, TRANSPORT) and deployment complexity that raise the overall setup burden. Per the static calibration rule, setup is capped at 15 absent verifiable CI/test evidence, so it is scored at 13.

Static review · not runListed 2026-09-07

Read the FMRS scoring method →

Fit and risk

What it can accessReads local filesUses the networkChanges third-party account data

Best for

  • Bioinformatics researchers who want AI assistants to operate the Galaxy platform
  • Galaxy instance administrators and regular Galaxy users
  • Teams automating analysis workflows

Not for

  • Users not on the Galaxy platform
  • Environments without a Galaxy URL and API key
  • General-purpose bioinformatics outside Galaxy

Required permissions

  • Requires the target Galaxy instance URL (GALAXY_URL, required)
  • Requires a Galaxy API key (GALAXY_API_KEY, required and secret; may be omitted when using OAuth)
  • OAuth mode requires a session secret (GALAXY_MCP_SESSION_SECRET) and public URL (GALAXY_MCP_PUBLIC_URL)

Risks and side effects

  • The API key is a sensitive credential; leakage grants others access to your Galaxy account and analyses
  • Tools and workflows executed by the AI consume Galaxy compute resources
  • Remote HTTP deployment exposes a network port and requires careful OAuth and session-secret configuration

Setup

Before you start

Runtime：Python · Docker · Node.js

GALAXY_URL required URL of the Galaxy instance to connect to (e.g., https://usegalaxy.org); required.
GALAXY_API_KEY requiredsecret API key for authenticating with the Galaxy instance, generated in your Galaxy user preferences; required (can be omitted when using OAuth).
GALAXY_MCP_SESSION_SECRET optionalsecret Session secret for OAuth/HTTP mode; generate with openssl rand -hex 32.
Other optional settings (2)
GALAXY_MCP_PUBLIC_URL optional Public base URL used for OAuth redirects when running HTTP transport with OAuth.
GALAXY_MCP_TRANSPORT optional Transport setting (e.g., streamable-http), used in container deployments to switch transport.

Run via uvx (uvx galaxy-mcp, stdio by default), or pip install galaxy-mcp and run galaxy-mcp, or use the Docker image galaxyproject/galaxy-mcp. Set required environment variables GALAXY_URL (Galaxy instance URL) and GALAXY_API_KEY (API key). For HTTP: uvx galaxy-mcp --transport streamable-http --host 0.0.0.0 --port 8000. For Claude Desktop, add a galaxy-mcp entry to claude_desktop_config..

claude_desktop_config.json 
{"mcpServers":{"galaxy-mcp":{"command":"uvx","args":["galaxy-mcp"],"env":{"GALAXY_URL":"https://usegalaxy.org","GALAXY_API_KEY":"SECRETS"}}}}

Shown for Claude Desktop. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json
{
  "servers": {
    "galaxy-mcp": {
      "command": "uvx",
      "args": [
        "galaxy-mcp"
      ],
      "env": {
        "GALAXY_URL": "https://usegalaxy.org",
        "GALAXY_API_KEY": "SECRETS"
      }
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal
claude mcp add galaxy-mcp -e GALAXY_URL=https://usegalaxy.org -e GALAXY_API_KEY=SECRETS -- uvx galaxy-mcp

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

Check that galaxy-mcp appears as running in Claude Desktop's developer menu, then prompt the AI with "can you connect to galaxy"; if env vars are unset it will ask for connection details, and once connected it should be able to list your Galaxy server information.

Troubleshooting

  1. Verify GALAXY_URL and GALAXY_API_KEY are set correctly and point at the intended Galaxy instance
  2. Check the developer menu in Claude Desktop shows galaxy-mcp running; restart Claude Desktop if needed
  3. If env config is missing, provide connection details in chat (e.g., your Galaxy API key)
  4. For OAuth login, ensure GALAXY_MCP_PUBLIC_URL and GALAXY_MCP_SESSION_SECRET are configured

Things to try

Once connected, you can ask your AI assistant things like:

  • Can you connect to Galaxy?
  • Give a summary with my histories
  • Search Galaxy tools for RNA sequencing analysis
  • Import a workflow from the IWC to my Galaxy instance

Use cases

Search and execute Galaxy tools via conversation
Import and manage workflows from the IWC catalog
Manage Galaxy histories and datasets
Upload local files to Galaxy
Share one server across users with per-session credential isolation

Supported clients

Claude Desktop

Listed from the project's documentation, not tested by this site.

Overview

Galaxy MCP Server is an open-source (MIT) MCP server maintained by the Galaxy Project. It lets AI assistants connect to any Galaxy instance to search and execute tools, manage workflows and histories, upload files, and optionally authenticate via browser-based OAuth. The Python implementation is built on BioBlend, and the repository also ships a TypeScript workspace (galaxy-agent-tools) offering a galaxy-cli tool and a Node MCP server kept in lockstep.

Similar servers

Context7 80 · B

Upstash's official server providing up-to-date third-party library docs for AI coding assistants

★ 63.3k · Tools 2 Compare with this →
Compare with other MCP servers

Source revision 81a362569c7b Data synced 2026-10-19 Read the FMRS scoring method