Skip to content
MCP server

gitscrum-core

MCP server giving AI assistants access to GitScrum tasks, sprints, time tracking

Runs locally
Install

gitscrum-core is an MCP server for productivity. It runs locally on your machine and is started by your AI client.

Platforms
API, CLI, VS Code
Runs
Locally
Version
1.0.1
Listing updated

Similar MCP servers

Install

Run in a terminal

claude mcp add mcp-server --env GITSCRUM_TOKEN='<GITSCRUM_TOKEN>' -- npx -y @gitscrum-studio/mcp-server

Add to claude_desktop_config.json

{
  "mcpServers": {
    "mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@gitscrum-studio/mcp-server"
      ],
      "env": {
        "GITSCRUM_TOKEN": "<GITSCRUM_TOKEN>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@gitscrum-studio/mcp-server"
      ],
      "env": {
        "GITSCRUM_TOKEN": "<GITSCRUM_TOKEN>"
      }
    }
  }
}

Add to .vscode/mcp.json

{
  "servers": {
    "mcp-server": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@gitscrum-studio/mcp-server"
      ],
      "env": {
        "GITSCRUM_TOKEN": "<GITSCRUM_TOKEN>"
      }
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@gitscrum-studio/mcp-server"
      ],
      "env": {
        "GITSCRUM_TOKEN": "<GITSCRUM_TOKEN>"
      }
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@gitscrum-studio/mcp-server"
      ],
      "env": {
        "GITSCRUM_TOKEN": "<GITSCRUM_TOKEN>"
      }
    }
  }
}

Using several servers? Build one config file with the MCP config generator.

Configuration

  • GITSCRUM_API_URLGitScrum API base URL
  • GITSCRUM_TOKENGitScrum API authentication token

About gitscrum-core

The README says this server turns an AI assistant into a project manager for a GitScrum workspace. It connects clients such as Claude, Cursor, Windsurf and GitHub Copilot to tasks, sprints, user stories, epics, kanban workflows, team discussions, wiki and notes, and time tracking through the Model Context Protocol.

Beyond core project management, the tool set covers a ClientFlow CRM area with clients, invoices and proposals, plus Insights reports for standups, analytics, activity feeds and budget tracking. Tools use a consolidated action parameter, so one tool name covers several operations.

According to the README, the server allows create, read and update operations only, with delete blocked at both the MCP and API layers. Authentication uses an OAuth 2.0 device code flow. It can be run locally over stdio through npm or used through a hosted SSE endpoint. It suits GitScrum users who want to manage their workspace from an AI chat.

Tools this server provides (5)

Taken from the project documentation; the running server may differ.

task
Task actions including my, today, get, create, update, complete and move.
sprint
Sprint actions including list, create, update, stats and progress.
time
Time tracking actions including active, start, stop, logs and analytics.
standup
Standup actions including summary, completed, blockers and digest.
budget
Budget actions including projects_at_risk, overview and burn_down.

What it does

Task and sprint control
Lets an assistant list, create, update and complete tasks and manage sprints within a GitScrum workspace.
Time tracking
Starts and stops timers and pulls logs, reports and team productivity data.
Client CRM and invoicing
Covers clients, invoices and proposals under the ClientFlow module.
Standup and budget reports
Generates standup digests, blocker lists and budget burn-down data.
Device code sign-in
Authenticates through an OAuth 2.0 device authorization flow in the browser, so credentials are not passed to the server.
Read and write only
The README states that delete operations are blocked at the MCP and API layers.

Good for

  • Asking an AI assistant what tasks are due today across projects
  • Creating a sprint and assigning backlog items through chat
  • Checking which projects are over budget or at risk
  • Generating a team standup digest from completed work and blockers
  • Drafting and sending client proposals from an AI client

gitscrum-core FAQ

How do I connect gitscrum-core to my AI client?

The README gives an npx command (@gitscrum-studio/mcp-server) for a local stdio setup, with config examples for Claude Desktop and VS Code or Cursor. A hosted SSE endpoint is also listed.

Can gitscrum-core delete data in GitScrum?

No. The README says only create, read and update operations are allowed, and destructive operations must be done in the GitScrum Studio web app.

What are the requirements to run gitscrum-core locally?

The README lists Node.js 18 or later and npm 8 or later.

What license is gitscrum-core released under?

The README states the project is released under the MIT license.

How do I add gitscrum-core to Claude Code?

Run this command in your terminal: claude mcp add mcp-server --env GITSCRUM_TOKEN='<GITSCRUM_TOKEN>' -- npx -y @gitscrum-studio/mcp-server

What are the best alternatives to gitscrum-core?

The closest alternatives to gitscrum-core are Orangescrum MCP Server, Monday Com, IterDone, Tack and Microsoft Todo MCP Server.

Where this comes from

  • Description: written from the project's README and its MCP Registry entry.
  • Install steps: from the official MCP Registry entry.

How AIPoweredBot checks servers