Model Context Protocol (MCP) server that connects Claude with Google Tasks, exposing tools for listing, creating, updating and deleting task-lists and tasks.
https://github.com/arpitbatra123/mcp-googletasksTired of context switching between Claude and Google Tasks during your workflow? This MCP server brings your entire Google Tasks ecosystem directly into Claude conversations, turning your AI assistant into a full-featured task management interface.
Every time you're deep in a Claude conversation about a project and need to create tasks, check deadlines, or update task status, you're paying the "app switching tax" - breaking your flow to jump between tools. This MCP server eliminates that friction entirely.
You can now:
Project Planning Sessions: Instead of taking notes in Claude then manually transferring to Google Tasks, create your entire project structure in real-time during the conversation. "Create a 'Q1 Launch' task list with tasks for design review, development, testing, and deployment" - done.
Daily Standup Prep: Ask Claude to pull your tasks for today across all lists, analyze what's overdue, and help prioritize your work without opening Google Tasks once.
Meeting Follow-ups: During project discussions with Claude, instantly convert action items into properly categorized tasks: "Add these three deliverables to the client project list with next Friday's deadline."
Context-Aware Task Management: Claude can analyze your task patterns, suggest reorganization, and help you break down complex tasks into manageable subtasks - all while maintaining access to your actual Google Tasks data.
The authentication flow is straightforward OAuth2 - you'll authenticate once per session through your browser, then Claude has full access to your Google Tasks ecosystem. The server provides complete CRUD operations for both task lists and individual tasks.
What sets this apart is the comprehensive API coverage: you're not limited to basic create/read operations. You can move tasks between lists, update due dates, mark tasks complete, and even bulk-clear completed tasks - everything you'd do in the Google Tasks interface, but through natural language in Claude.
The TypeScript implementation includes proper error handling and follows Google's API best practices. You'll need a Google Cloud project with Tasks API enabled and OAuth credentials, but the setup process is well-documented.
Install via Smithery for automated setup:
npx -y @smithery/cli install @arpitbatra123/mcp-googletasks --client claude
Or manual installation with your own Google Cloud credentials for more control over the configuration.
This particularly shines for developers who use Google Tasks to track feature development, bug fixes, and project milestones. You can have Claude analyze your codebase, suggest task breakdowns, and immediately create the corresponding tasks in your project management system.
The server maintains session-based authentication, so you'll need to re-authenticate when you restart, but that's a reasonable trade-off for the security model.
Whether you're planning sprints, tracking deliverables, or just trying to keep your personal tasks organized while working with Claude, this MCP server turns your AI conversations into a complete task management workflow.