A Model Context Protocol (MCP) server that integrates with Google Tasks, providing tools to list, search, create, update, delete and clear tasks.
https://github.com/zcaceres/gtasks-mcpStop alt-tabbing between your AI assistant and Google Tasks. This MCP server bridges the gap, letting you manage your entire task workflow without breaking your conversational flow.
You're deep in a conversation with Claude, planning your next sprint or brainstorming features. A task comes up. You switch to Google Tasks, create it, maybe add some notes, then switch back to continue the conversation. Rinse and repeat.
Each context switch costs you 15-20 seconds of refocus time. Multiply that by the dozens of tasks you create daily, and you're losing serious productivity to app switching.
This MCP server gives your AI assistant direct access to Google Tasks with six powerful operations:
Smart Task Creation: "Create a task to refactor the authentication module, due next Friday with notes about the security requirements we discussed." Done. No switching, no copy-pasting context.
Intelligent Search: "Find all tasks related to the payment system." Your assistant searches across titles and notes instantly, surfacing exactly what you need.
Bulk Operations: "Clear all completed tasks from my development project list." One command handles what would take multiple clicks in the UI.
Context-Aware Updates: "Update that API integration task to completed and add notes about the performance improvements." Your assistant knows which task you're referring to from your conversation history.
Sprint Planning: "List all tasks in my current sprint, then create three new tasks for the user authentication epic we just discussed." Your assistant becomes your project manager.
Code Review Workflows: "Create a task to address the security concerns in PR #142, assign it to the security review list, due tomorrow." Task creation happens in the same conversation where you're discussing the code.
Bug Triage: "Search for any existing tasks about the payment gateway timeout issue." No more duplicate tasks because your assistant can check what already exists.
Daily Standups: "Show me all tasks assigned to me that are due this week, then create a follow-up task for the database migration discussion." Your standup prep happens conversationally.
Natural Language Processing: Instead of clicking through forms, describe what you want. "Create a high-priority task about optimizing the database queries with a due date of next Tuesday."
Conversation Context: Your assistant remembers what you've been discussing. Reference "that API issue we talked about" and it knows which task you mean.
Batch Operations: "Create five tasks for the different microservices we need to containerize." Your assistant handles the repetitive work.
Cross-Reference Power: "Find all tasks related to the payment system and create a summary task for the security audit." Connections your assistant can make that you might miss.
The Google Cloud Console setup is standard OAuth fare - create a project, enable the Tasks API, configure credentials. The server handles the authentication flow automatically.
One-liner installation via Smithery:
npx -y @smithery/cli install @zcaceres/gtasks --client claude
Authentication is a one-time browser flow. After that, your assistant has persistent access to your tasks with proper OAuth token management.
Add this to your Claude Desktop configuration:
{
"mcpServers": {
"gtasks": {
"command": "/opt/homebrew/bin/node",
"args": ["{ABSOLUTE_PATH}/dist/index.js"]
}
}
}
No workflow changes required. Your assistant simply gains task management capabilities. You can still use Google Tasks normally - the MCP server works alongside, not instead of, the regular UI.
This isn't just about saving clicks. It's about maintaining mental flow state while ensuring nothing falls through the cracks. Your planning conversations become actionable without interruption. Your task management becomes as fluid as your thinking.
When your AI assistant can create, search, and manage tasks conversationally, task management stops being a separate activity and becomes part of your natural workflow.
Time to give your AI assistant some real productivity powers.