Skip to main content
Glama
cnavta

Sprint MCP Server

by cnavta

Sprint MCP Server

A combination agent workflow and MCP server focused around iterative, collaborative development between Humans and Agents.

Overview

This MCP server implements tools that support the Sprint Protocol defined in AGENTS.md. It enables LLM agents (like Claude) to manage structured sprint-based development workflows through the Model Context Protocol.

Related MCP server: Project Manager MCP

Features

  • Sprint Management: Initialize, track, and complete sprints following the Sprint Protocol

  • Status Checking: Verify active sprints and enforce single-sprint rule (S3)

  • Manifest Generation: Automatically create sprint manifests with proper metadata

  • Request Logging: Track all actions and decisions in request logs

  • Protocol Enforcement: Validate sprint lifecycle rules and requirements

Installation

Prerequisites

  • Node.js >= 18.0.0

  • npm or yarn

Setup

  1. Clone this repository

  2. Install dependencies:

npm install
  1. Build the project:

npm run build

Usage

Running the Server

The MCP server communicates via stdio:

npm run dev

Or run the built version:

node dist/index.js

Connecting to Claude Desktop

Add this server to your Claude Desktop configuration (~/Library/Application Support/Claude/claude_desktop_config.json on macOS):

{
  "mcpServers": {
    "sprint-mcp": {
      "command": "node",
      "args": ["/path/to/sprint-mcp/dist/index.js"]
    }
  }
}

After restarting Claude Desktop, the sprint tools will be available.

Available Tools

start-sprint

Initialize a new sprint with manifest and directory structure.

Parameters:

  • title (string, required): Concise sprint title

  • goal (string, required): Clear sprint objective

  • owner (string, required): GitHub handle or name of sprint owner

Example:

{
  "title": "Implement User Profile Service",
  "goal": "Create microservice for user profile management with REST API",
  "owner": "@johndoe"
}

Behavior:

  • Checks for active sprints (rule S3)

  • Generates unique sprint ID: sprint-<number>-<hash>

  • Creates sprint directory in planning/

  • Generates sprint-manifest.yaml and request-log.md

  • Suggests feature branch name

  • Sets status to planning

check-sprint-status

Verify current sprint state and check for active sprints.

Parameters: None

Returns:

  • List of active sprints with details

  • Count of completed sprints

  • Warning if multiple active sprints detected (protocol violation)

  • Confirmation if ready to start new sprint

Development

Build

npm run build

Watch Mode

npm run watch

Testing

npm test

Run tests in watch mode:

npm run test:watch

Generate coverage report:

npm run test:coverage

Project Structure

sprint-mcp/
├── src/
│   ├── index.ts              # MCP server entry point
│   ├── common/
│   │   ├── logger.ts          # Logging facade
│   │   └── file-utils.ts      # File system utilities
│   ├── tools/
│   │   ├── start-sprint.ts    # Start sprint tool implementation
│   │   └── check-sprint-status.ts  # Status check tool
│   └── types/
│       └── sprint.ts          # TypeScript type definitions
├── planning/                   # Sprint artifacts directory
├── architecture.yaml          # Canonical source of truth
├── AGENTS.md                  # Sprint Protocol definition
├── CLAUDE.md                  # Claude Code guidance
└── package.json

Sprint Protocol

This server implements the Sprint Protocol defined in AGENTS.md. Key principles:

  1. Precedence: architecture.yaml > AGENTS.md > everything else

  2. Sprint Control: Only one sprint active at a time (rule S3)

  3. Lifecycle: Plan → Approve → Implement → Validate → Verify → Publish → Retro → Learn

  4. Traceability: All actions logged in request-log.md

  5. Definition of Done: Code quality, testing, deployment, documentation, traceability

Logging

Logs are written to stderr to avoid interfering with MCP stdio protocol.

Set log level via environment variable:

LOG_LEVEL=debug node dist/index.js

Available levels: debug, info, warn, error

License

MIT

F
license - not found
-
quality - not tested
B
maintenance

Maintenance

Maintainers
Response time
Release cycle
Releases (12mo)
Commit activity

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Servers

View all related MCP servers

Related MCP Connectors

  • The project brain for AI coding agents — memory, decisions, sprints, knowledge base via MCP.

  • The team layer for AI coding agents: shared contracts, collision alerts, E2EE sessions.

  • Cross-agent artifact workspace with provenance across Claude Code, Codex, Cursor, LangGraph.

View all MCP Connectors

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/cnavta/sprint-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server