🚀 Taskade MCP Agent

Manage your entire Taskade workspace with plain English

An AI-powered workspace agent that connects to Taskade via the Model Context Protocol (MCP) — create projects, organize tasks, and automate workflows using natural language.

Live Demo · Features · How It Works · Quick Start · Usage · Deployment

---

💡 Why This Project?

Managing projects shouldn't feel like a project itself. Instead of clicking through menus to create tasks, check deadlines, or reorganize workflows, what if you could just say what you want?

"Show me all incomplete tasks across my projects" "Add a task called 'Review PR' to my Sprint board" "What's due this week?"

Taskade MCP Agent turns those sentences into real actions in your Taskade workspace. It combines a modern LLM (Qwen3-30B-A3B via Nebius AI) with Taskade's official MCP server, so the AI doesn't guess about your workspace — it talks to the Taskade API through a secure, structured protocol.

✨ Features

• 📁 Project Management — Create, list, and manage Taskade projects and workspaces conversationally
• ✅ Task Operations — Add, update, complete, and organize tasks across projects
• 🤖 AI Agent Integration — Interact with Taskade's autonomous AI agents
• 💬 Natural Language Interface — No syntax to learn; just describe what you want
• 🔌 MCP Protocol — Uses Taskade's official MCP server for secure, structured API access
• 🎛️ Query Templates — One-click templates for common operations (list projects, create tasks, manage workflows)
• 🔐 Bring Your Own Keys — API keys are entered at runtime in the sidebar; nothing is hardcoded

🧠 How It Works

┌─────────────┐     ┌──────────────┐     ┌─────────────────────┐     ┌─────────────┐
│   You type   │     │  Agno Agent  │     │  Taskade MCP Server │     │   Taskade   │
│  a request   │ ──▶ │ (Qwen3-30B   │ ──▶ │  (@taskade/mcp-     │ ──▶ │     API     │
│ in Streamlit │     │  via Nebius) │     │   server via npx)   │     │             │
└─────────────┘     └──────────────┘     └─────────────────────┘     └─────────────┘
        ▲                                                                    │
        └────────────────── formatted results ◀──────────────────────────────┘

1. You ask — type a natural-language query in the Streamlit UI
2. The agent reasons — an Agno agent powered by Qwen3-30B-A3B (Nebius AI) decides which Taskade tools to call
3. MCP executes — the official Taskade MCP server translates tool calls into authenticated Taskade API requests
4. You get results — clean, markdown-formatted answers with tables, statuses, and due dates

🛠️ Tech Stack

Component	Technology
Language	Python 3.10+
Web UI	Streamlit
Agent Framework	Agno
LLM Provider	Nebius AI — Qwen3-30B-A3B
Tool Protocol	Taskade MCP Server (Node.js)
Config	python-dotenv

📋 Prerequisites

• Python 3.10+
• Node.js 18+ (runs the Taskade MCP server via npx)
• A Taskade account and API key — generate one in Taskade Settings
• A Nebius AI API key

⚡ Quick Start

1. Clone the repository

git clone https://github.com/tirth1263/taskade-mcp-agent.git
cd taskade-mcp-agent

2. Create and activate a virtual environment

python -m venv .venv
source .venv/bin/activate   # On Windows: .venv\Scripts\activate

3. Install dependencies

# Using uv (recommended)
uv sync

# Or using pip
pip install -r requirements.txt

4. Install the Taskade MCP server

npm install -g @taskade/mcp-server

5. Configure environment variables

Create a .env file in the project root (or copy .env.example):

NEBIUS_API_KEY="your_nebius_api_key"
TASKADE_API_KEY="your_taskade_api_key"

💡 You can also skip the .env file and enter both keys directly in the app's sidebar.

🚀 Usage

1. Run the application

streamlit run main.py

2. Open your browser at http://localhost:8501

3. Configure — in the sidebar, enter your Nebius API key and Taskade API key, then click Save Configuration

4. Start asking — use the query input to interact with your Taskade workspace:

Intent	Example Query
📂 List Projects	"Show all my projects"
➕ Create Tasks	"Add a task called 'Review PR' to my project"
🗓️ Manage Workflows	"What tasks are due this week?"
🔍 Custom Queries	Ask anything about your Taskade workspace

☁️ Deploy Your Own

This repo is deployment-ready for Streamlit Community Cloud — free hosting, no server required:

1. Fork this repository
2. Go to share.streamlit.io and sign in with GitHub
3. Click Create app → pick your fork → set Main file path to main.py → Deploy

The included packages.txt automatically installs Node.js and npm on the deployment container, so the Taskade MCP server (npx -y @taskade/mcp-server) works out of the box. Visitors bring their own API keys via the sidebar — no secrets are stored in the deployment.

📁 Project Structure

taskade-mcp-agent/
├── .streamlit/
│   └── config.toml      # Streamlit theme configuration
├── .env.example         # Example environment variables
├── main.py              # Streamlit application & MCP agent
├── packages.txt         # System packages for Streamlit Cloud (Node.js)
├── requirements.txt     # Python dependencies
├── LICENSE              # MIT license
└── README.md            # You are here 📍

🌐 About Taskade

Taskade is an AI-native workspace platform for building apps, deploying autonomous AI agents, and automating workflows with 100+ integrations:

• 🤖 Taskade AI Agents — deploy autonomous agents that work across your projects
• 🧬 Taskade Genesis — build custom AI-powered apps visually
• 🔌 Taskade MCP Server — connect AI tools to Taskade via the Model Context Protocol
• ⚙️ Workflow Automation — automate repetitive tasks with AI-powered workflows

🙏 Acknowledgments

• Based on the taskade_mcp_agent example from Arindam200/awesome-ai-apps
• Taskade for the official MCP server
• Agno for the agent framework
• Nebius AI for LLM inference

📄 License

This project is licensed under the MIT License — see LICENSE for details.

---

⭐ If this project helped you, consider giving it a star!

Made with ❤️ by Tirth Rank