Skip to main content
Glama
globus-mcp
by globus

Globus MCP Server

The Globus MCP Server enables LLM applications to interact with Globus services.

Supported Tools

Globus Transfer

  • globus_transfer_list_endpoints_and_collections - List endpoints and collections the user has access to

  • globus_transfer_search_endpoints_and_collections - Use a filter string to search all endpoints and collections that are visible to the user

  • globus_transfer_submit_task - Submit a transfer task between two collections

  • globus_transfer_get_task_events - Get a list of task events

  • globus_transfer_list_directory - List contents of a directory on a collection

Globus Compute

  • globus_compute_list_endpoints - List endpoints that the user has access to

  • globus_compute_register_python_function - Register a Python function

  • globus_compute_register_shell_command - Register a shell command

  • globus_compute_submit_task - Submit a task to an endpoint

  • globus_compute_get_task_status - Retrieve the status and result of a task

Related MCP server: jlab-mcp

Configuration

The following configuration is compatible with most LLM applications that support MCP such as Claude Desktop:

{
  "mcpServers": {
    "globus-mcp": {
      "command": "uvx",
      "args": ["globus-mcp"]
    }
  }
}

Limiting Tool Registration

By default, the Globus MCP server registers tools for every service. To register tools for only specific services, use the --services command-line flag:

{
  "mcpServers": {
    "globus-mcp": {
      "command": "uvx",
      "args": [
        "globus-mcp",
        "--services",
        "compute"
      ]
    }
  }
}

Specifying Client Credentials

If you've registered a client application in the Globus web UI, you can specify the client credentials via the GLOBUS_CLIENT_ID and GLOBUS_CLIENT_SECRET environment variables:

{
  "mcpServers": {
    "globus-mcp": {
      "command": "uvx",
      "args": ["globus-mcp"],
      "env": {
        "GLOBUS_CLIENT_ID": "...",
        "GLOBUS_CLIENT_SECRET": "..."
      }
    }
  }
}

Available Tools

10 tools
globus_compute_get_task_statusB

Retrieve the status and result of a Globus Compute task.

Inputs: task_id (required). Returns: result, status, task_id, exception

globus_compute_list_endpointsB

List Globus Compute endpoints that the user has access to.

Inputs: role (required). Returns: result

globus_compute_register_python_functionA

Register a Python function with Globus Compute.

Use globus_compute_submit_task to run the registered Python function on an endpoint.

Inputs: public (required), description (required), function_code (required), function_name (required). Returns: function_id

globus_compute_register_shell_commandA

Register a shell command function with Globus Compute.

Use globus_compute_submit_task to run the registered shell command on an endpoint.

Inputs: public (required), command (required), timeout (required), description (required). Returns: function_id

globus_compute_submit_taskA

Submit a function execution task to a Globus Compute endpoint.

Use globus_compute_get_task_status to monitor progress and retrieve results.

Inputs: endpoint_id (required), function_id (required), function_args (required), function_kwargs (required). Returns: task_id

globus_transfer_get_task_eventsA

Get a list of Globus Transfer task events to monitor the status and progress of a task. The events are ordered by time descending (newest first).

Inputs: limit (required), offset (required), task_id (required). Returns: data, limit, offset

globus_transfer_list_directoryB

List contents of a directory on a Globus Transfer collection

Inputs: path (required), limit (required), offset (required), collection_id (required). Returns: data, limit, offset

globus_transfer_list_endpoints_and_collectionsC

List Globus Transfer endpoints and collections that the user has access to, filtered based on the provided scope.

Inputs: limit (required), offset (required), filter_scope (required). Returns: data, limit, offset, has_next_page

globus_transfer_search_endpoints_and_collectionsA

Use a filter string to search all Globus Transfer endpoints and collections that are visible to the user.

Inputs: limit (required), offset (required), filter_fulltext (required). Returns: data, limit, offset, has_next_page

globus_transfer_submit_taskA

Submit a transfer task between two Globus Transfer collections.

Use globus_transfer_get_task_events to monitor the task's progress.

Inputs: label (required), source_path (required), destination_path (required), source_collection_id (required), +1 more. Returns: task_id

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.7/5.0

    Scored across 10 tools

    Disambiguation5/5

    The tools are cleanly separated into two distinct domains (Globus Compute and Globus Transfer), with no overlap in functionality. Each tool has a unique and clear purpose, aided by detailed descriptions that specify when to use each.

    Naming Consistency5/5

    All tool names follow a consistent pattern: globus_{domain}_{verb}_{noun} in snake_case. This makes it predictable and easy for an agent to infer functionality from the name.

    Tool Count5/5

    With 10 tools covering the key operations for both Compute and Transfer services, the count is well-scoped. Each tool addresses a distinct need without unnecessary duplication, and the set feels complete for typical workflows.

    Completeness4/5

    The tool surface covers the essential operations for both services: registration, listing, submission, and status retrieval for Compute; listing, searching, directory browsing, submission, and event monitoring for Transfer. Minor omissions like delete or cancel are present but don't significantly hinder core usage.

    Maintenance

    ActivityMaintained
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • F
      license
      A
      quality
      C
      maintenance
      An MCP server that enables LLMs to execute Python code on GPU-accelerated compute nodes within SLURM-managed HPC environments. It bridges local clients to remote clusters by launching JupyterLab sessions via SLURM jobs to facilitate high-performance notebook-based computation.
      7
      -
    • A
      license
      Not graded
      quality
      B
      maintenance
      An MCP server that exposes Rucio distributed data management operations as tools for LLMs. Designed for ATLAS physicists working with grid data on analysis facilities, but usable with any Rucio instance.
      6
      Apache 2.0