MCP 服务器

MCP-GET

A unified Model Context Protocol server that provides a consistent interface for AI assistants to interact with productivity tools like Linear, GitHub, Slack, and Notion. It enables users to search, retrieve, and manage tasks and data across multiple workplace services from a single endpoint.

README

MCP-GET

A flexible Model Context Protocol server for accessing multiple task management and productivity services. MCP-GET provides a unified interface for AI assistants to interact with popular workplace tools through a consistent, extensible architecture.

About the Model Context Protocol

The Model Context Protocol (MCP) is an open standard that enables AI assistants to securely connect to external data sources and tools. Developed by Anthropic, MCP provides a standardized way for AI systems to access real-time information and perform actions across different services.

Key Resources:

Setup

Install dependencies:
```
npm install
```

Configure your services in the .env file:

# Linear (currently supported)
# Get your API key: https://linear.app/settings/api
LINEAR_API_KEY=your_linear_api_key_here
LINEAR_ENABLED=true

# Notion
# Create integration: https://developers.notion.com/docs/create-a-notion-integration
NOTION_API_KEY=your_notion_integration_token
NOTION_ENABLED=true

# Slack
# Create app: https://api.slack.com/apps
SLACK_BOT_TOKEN=xoxb-your-slack-bot-token
SLACK_ENABLED=true

# GitHub
# Generate token: https://github.com/settings/tokens
GITHUB_API_KEY=your_github_personal_access_token
GITHUB_ENABLED=true

# Plaid
# Get credentials: https://dashboard.plaid.com/developers/keys
PLAID_CLIENT_ID=your_plaid_client_id
PLAID_SECRET=your_plaid_secret_key
PLAID_ENV=sandbox  # or development, production
PLAID_ENABLED=true

Run the server:
```
npm start
```

Available Tools

Linear Service

linear_get_my_issues: Get recent issues assigned to you from Linear
linear_get_issue_details: Get detailed information about a specific Linear issue
linear_search_issues: Search for Linear issues by title or description

Notion Service (Planned)

notion_get_my_pages: Get recent pages you have access to
notion_get_page_details: Get detailed information about a specific page
notion_search_pages: Search for pages by title or content

Slack Service (Planned)

slack_get_my_messages: Get recent messages from your conversations
slack_get_thread_details: Get detailed information about a specific thread
slack_search_messages: Search for messages by content or user

GitHub Service (Planned)

github_get_my_issues: Get recent issues assigned to you
github_get_issue_details: Get detailed information about a specific issue
github_search_issues: Search for issues by title or description

Plaid Service (Planned)

plaid_get_accounts: Get linked bank accounts and balances
plaid_get_transactions: Get recent transactions across accounts
plaid_search_transactions: Search transactions by merchant, amount, or description

Architecture

MCP-GET uses a service-based architecture that makes it easy to add new productivity services:

mcp-get/
├── index.js              # Main MCP server
├── config.js             # Service configuration
└── services/
    ├── base.js           # Base service class
    ├── linear.js         # Linear service implementation
    ├── index.js          # Service registry
    └── [future services] # Jira, GitHub, etc.

Each service implements the same interface:

getMyIssues(limit): Get user's assigned items
getIssueDetails(id): Get detailed item information
searchIssues(query, limit): Search items by text

Integration with AI Assistants

MCP-GET integrates seamlessly with AI assistants that support the Model Context Protocol:

Claude Code

Claude Code is Anthropic's official CLI tool that supports MCP servers. Add MCP-GET to your Claude Code configuration to access multiple productivity services directly in your conversations.

Claude Code Resources:

Configuration Example

{
  "mcpServers": {
    "mcp-get": {
      "command": "node",
      "args": ["path/to/mcp-get/index.js"],
      "env": {
        "LINEAR_API_KEY": "your_linear_api_key",
        "GITHUB_API_KEY": "your_github_token",
        "NOTION_API_KEY": "your_notion_integration_token",
        "SLACK_BOT_TOKEN": "your_slack_bot_token",
        "PLAID_CLIENT_ID": "your_plaid_client_id",
        "PLAID_SECRET": "your_plaid_secret"
      }
    }
  }
}

Other Compatible AI Assistants

Cline - VS Code extension with MCP support
Continue - AI coding assistant with MCP integration
Custom implementations using the MCP SDK

Benefits

Unified Interface: Consistent command patterns across all services
Contextual Awareness: AI assistants can cross-reference data between services
Productivity Enhancement: Streamlined workflows without switching between tools
Extensible Architecture: Easy to add new services as your workflow evolves

Adding New Services

Create a new service file in services/ (e.g., jira.js)
Extend the BaseService class
Implement the required methods
Add the service to the registry in services/index.js
Add configuration in config.js

See services/linear.js for a complete example.

Development Resources:

Supported Services

Currently Available:

Linear: Full support for issues, assignments, and search functionality

Planned Integration:

Notion: Database queries, page management, and content search
Slack: Message retrieval, thread navigation, and workspace search
GitHub: Issues, pull requests, and repository management
Plaid: Banking data access, transaction history, and account management

Development Roadmap

MCP-GET is designed to become the definitive MCP server for workplace productivity tools. Our development roadmap includes:

Phase 1: Core Services (Current)

Linear integration (Complete)
GitHub Issues and Pull Requests
Plaid financial data access

Phase 2: Communication Platforms

Slack workspace integration
Microsoft Teams connectivity
Discord server management

Phase 3: Knowledge Management

Notion database and page access
Confluence space management
Obsidian vault integration

Phase 4: Financial & Analytics

Enhanced Plaid transaction analysis
Banking workflow automation
Financial reporting and insights

Contributing

We welcome contributions to expand service support and improve functionality. To add support for a new service:

Create a service class extending BaseService in the services/ directory
Implement required methods following the established interface pattern
Add configuration options to config.js with appropriate environment variables
Update the service registry in services/index.js
Add comprehensive tests and update documentation