Whilst the race of making products / services AI-compatible, Zerodha unleashes Kite MCP, allowing users to automate the trading processes in many possible ways including getting AI-powered insights of your portfolio.
In the fast-paced world of stock markets, every advantage counts. For years, active traders and developers have leveraged APIs to automate aspects of their investment strategies. Now, Zerodha, India’s largest stockbroker, has taken a significant leap forward by introducing Kite MCP (Model Context Protocol). This innovative feature isn’t just another update; it’s a game-changer designed to automate and streamline your investing and trading process by bridging your Zerodha account directly with advanced AI assistants.
What is Kite MCP?
At its core, Kite MCP (Model Context Protocol) is an open standard that allows AI systems (like Large Language Models or LLMs) to securely connect with external tools and data sources in a standardized, “plug-and-play” manner. Zerodha’s Kite MCP specifically acts as a backend server program, seamlessly integrating with the existing Kite Connect API. Think of it as a universal translator, enabling your chosen AI assistant to understand and interact with your Zerodha trading account in real-time, using natural language.
Traditionally, using the Kite Connect API required coding skills to query market data, analyze portfolios, and manage trades. With Kite MCP, that complexity is drastically reduced. It provides context to the AI, feeding it current portfolio data, past trades, profit & loss (P&L) statements, and live market data, allowing the AI to “see” your information.
How Kite MCP Helps Users
Kite MCP fundamentally changes how you interact with your investments. It transforms your AI assistant into a personal financial advisor, capable of answering complex market questions and providing insights based on your specific portfolio. Here’s how it helps:
- Natural Language Queries: Instead of writing complex code or manually sifting through data, you can simply ask your AI assistant questions in plain English, like “What’s my P&L today?” or “Which sector dragged my portfolio down today?” The AI will then access your Zerodha data via MCP and provide personalized responses.
- Real-time Portfolio Analysis: Get instant insights into your holdings, positions, margins, and mutual fund investments. The AI can pull live market prices and even suggest rebalancing or optimization strategies.
- Automated Data Retrieval: Whether you need historical market data for a specific index or want to review your recent orders and trades, the AI can fetch this information swiftly and present it in an easily digestible format.
- Dynamic Actions (with permissions): While primarily read-only for security, Kite MCP can enable certain actions like managing Good Till Triggered (GTT) orders, depending on the AI client and explicit user permissions.
- Personalized Insights: Unlike generic market commentary, the AI leverages your actual portfolio data to offer tailored analysis, identifying top performers, underperformers, and sector exposure.
- Streamlined Workflow: For traders, it means no more switching between multiple apps or dashboards. Everything becomes conversational, making portfolio management less about searching and more about understanding.
Connecting Kite MCP with AI Platforms (e.g., Claude, Cursor)
Zerodha’s Kite MCP (Model Context Protocol) is a fantastic tool that allows you to connect your Zerodha trading account directly to AI assistants like Claude or Cursor. This means you can ask questions about your portfolio in plain English and get intelligent, personalized responses. While it might sound a bit technical at first, the process is quite straightforward. Here’s a step-by-step guide to get you connected:
What You’ll Need Before You Start
Before diving into the setup, make sure you have these prerequisites in place:
-
-
A Zerodha Trading Account: You need an active Zerodha trading account with Kite Connect API access. Good news: access to the Kite Connect API is typically free for Zerodha clients.
-
Node.js Installed: Kite MCP relies on Node.js (version 14 or above). If you don’t have it, download and install the LTS (Long Term Support) version from the official Node.js website (nodejs.org). You can verify your installation by opening your computer’s command prompt or terminal and typing
node --versionandnpm --version. -
Your Chosen AI Assistant Desktop App: This guide primarily focuses on connecting with AI desktop apps like Claude Desktop or Cursor AI, as they have built-in support for MCP. Make sure you have one of these installed on your computer.
-
Follow these steps carefully to bridge your Zerodha account with your AI assistant:
Step 1: Create a Zerodha Developer Account and App
Even though you’re not building a full app, Kite MCP uses the same secure API infrastructure.
-
-
Go to Kite Connect: Visit developers.kite.trade and sign up for a developer account. Log in to your newly created account.
-
Create a New App: Navigate to the “Apps” section in your Kite Developer dashboard. Click on “Create a new app.”
-
Fill in App Details:
-
App Name: Choose a descriptive name (e.g., “My AI Trading Assistant” or “Claude Zerodha Integration”).
-
App Category: Select “Personal” or an appropriate category.
-
Redirect URL: This is crucial. Set it exactly to
http://127.0.0.1:5000/zerodha/auth/redirect(or a similar address if your AI client specifies a different local port, though 5000 is common). -
Description: Briefly describe its purpose (e.g., “Connects my Zerodha account to AI assistant for portfolio analysis”).
-
-
Get API Credentials: After creating your app, you will see your API Key (also called Consumer Key) and API Secret (also called Consumer Secret). Keep these safe! You’ll need them in the next step.
-
Step 2: Configure Your AI Assistant for Kite MCP
This step involves telling your AI assistant how to “talk” to Zerodha via MCP. The exact steps vary slightly depending on your chosen AI tool.
For Claude Desktop:
-
-
Open Claude Settings: Launch the Claude Desktop application.
-
Access Developer Settings: Click the “Settings” (gear) icon, usually in the top-right corner. In the left sidebar, select “Developer.”
-
Edit Configuration File: Click on “Edit Config.” This will open a JSON configuration file (often
claude_desktop_config.json). -
Add Kite MCP Configuration: Carefully add or update the
mcpServersblock in this JSON file. Ensure you don’t accidentally delete existing content. Paste the following snippet:{ "mcpServers": { "kite": { "command": "npx", "args": ["mcp-remote", "https://mcp.kite.trade/sse"] } } }(Note: Some setups might require the
/mcpendpoint instead of/sse.https://mcp.kite.trade/mcpis also an option.) -
Save and Restart: Save the
claude_desktop_config.jsonfile and then completely restart your Claude Desktop application for the changes to take effect.
-
For Cursor AI (or VS Code with Copilot Chat):
-
-
Install Cursor CLI/Desktop Client: Make sure you have the Cursor desktop app installed.
-
Open Cursor Settings: Launch Cursor. Go to “Settings” (usually File > Preferences > Settings, or Ctrl+,).
-
Edit
settings.json: Search for “mcp” or “copilot chat mcp” (if using VS Code) and click “Edit in settings.json” if available. -
Add MCP Configuration: Add the following JSON snippet to your
settings.jsonfile:"mcp": { "inputs": [], "servers": { "kite": { "url": "https://mcp.kite.trade/sse" } } }(Again,https://mcp.kite.trade/mcpcan also be used if/ssedoesn't work.)
-
-
Save and Restart: Save the
settings.jsonfile and restart your Cursor (or VS Code) application.
Step 3: Authorize and Start Interacting
Now that your AI assistant knows about Zerodha, it’s time to connect your actual trading account.
-
-
Open a New Chat: In your AI assistant (e.g., Claude Desktop, Cursor), open a new chat session.
-
Look for MCP Tools: You might see a “Tools” or “Hammer” icon in the chat input area. Click it to confirm that “kite” (or “zerodha”) is listed as an available MCP server. If it’s there, you’re on the right track!
-
Initiate a Query: Type a simple query related to your Zerodha account, like:
-
“What is my current equity position?”
-
“Show me my total P&L today.”
-
“How much margin is available?”
-
-
Authorize Access (First Time Only): The first time you ask a Zerodha-related question, your AI assistant will prompt you to authorize access. It will provide a secure link.
-
Click the Link: This link will redirect you to Zerodha’s official login page.
-
Log In Securely: Log in using your Zerodha credentials and complete the 2-Factor Authentication (2FA). Zerodha emphasizes that your credentials are never stored by the AI client; the authentication happens securely through Zerodha’s system.
-
Grant Permissions: After successful login, you’ll be asked to grant permissions to the MCP app you created earlier. Grant the necessary permissions.
-
Return to AI: You’ll typically see a confirmation page. Go back to your AI assistant’s chat window.
-
-
Continue Your Query: Type “Done” or “Continue” (depending on the AI client) to signal that you’ve completed the authentication. Your AI assistant should now process your original query and provide a personalized response based on your Zerodha data!
-
What the Kite MCP can’t do? (as of now)
Here are five key demands of investors and traders that Zerodha’s Kite MCP currently doesn’t fully satisfy, despite its innovative features:
- Direct Automated Order Placement: While GTT orders can be managed, traders extensively need AI to place, modify, or cancel a wide range of complex orders (e.g., market, limit, stop-loss) automatically based on real-time market conditions without manual intervention.
- Advanced Algorithmic Trading Execution: Traders desire platforms for executing sophisticated algorithmic strategies directly via AI, including generating custom technical indicators and automatically initiating trades based on complex patterns. Kite MCP provides data but not the full execution framework for these custom algos.
- Proactive Real-Time Trade Alerts (Advanced): Beyond basic queries, investors seek proactive, automated alerts for specific price movements or portfolio changes, pushed dynamically to them without continuous querying, enabling immediate action.
- Integrated Scenario Planning with Live Trading: The ability for AI to simulate hypothetical trades and their impact on a live portfolio, providing dynamic projections and actionable steps, is a highly sought-after feature not fully covered.
- Consolidated Multi-Account Aggregation: Investors managing multiple trading accounts (personal, family) desire a holistic AI-driven view across all accounts for aggregated P&L, combined holdings analysis, and overall risk assessment, which is limited with the current single-account focus.
Pros and Cons of Kite MCP
Like any innovative technology, Kite MCP comes with its unique set of advantages and considerations:
| Pros | Cons |
| Enhanced Accessibility: Makes sophisticated portfolio analysis accessible to everyone, regardless of coding expertise, through natural language conversations. | Initial Setup Complexity: While not requiring advanced coding, the initial setup with Node.js and configuration files might be a slight hurdle for absolute non-technical users. |
| Significant Time Savings: Automates data retrieval and analysis, freeing up valuable time for traders and investors. | Read-Only Limitations: For most operations, Kite MCP is currently read-only. Direct order placement (buy/sell) is not yet broadly supported through conversational AI, though GTT management might be possible with specific permissions. Users cannot fully automate active trading strategies without additional custom coding. |
| Personalized Insights: Provides context-aware information and analysis directly from your Zerodha portfolio. | Reliance on AI Client Stability: The performance and stability can sometimes be dependent on the AI assistant’s own updates or potential temporary issues (as seen with some initial Claude integrations). |
| Security-Focused Design: Zerodha has implemented strict security measures, including read-only operations (for most data), non-storage of credentials by AI clients, and explicit user authorization. | Privacy Concerns (User Responsibility): While Zerodha’s implementation is secure, users still need to be mindful of the data shared with the AI assistant itself and understand the AI model’s own privacy policies. |
| Flexibility: Compatible with various leading AI assistants and code editors that support the Model Context Protocol. | |
| Cost-Effective: It’s a free addition for Zerodha users with Kite Connect API access, leveraging existing infrastructure. |
In conclusion, Zerodha’s Kite MCP is a significant stride towards making investing and trading more intuitive and automated. By seamlessly connecting your portfolio data with powerful AI assistants, it empowers users with instant, personalized insights, paving the way for a more efficient and data-driven approach to wealth management. It’s a clear indicator of how AI is set to redefine financial workflows.
Key Takeaways
- Kite MCP allows AI assistants to interact with your Zerodha account using natural language.
- It provides real-time portfolio analysis and automated data retrieval.
- Initial setup involves configuring your AI client with the MCP server URL.
- While powerful, it currently has limitations in direct automated order placement.
- Zerodha has implemented strict security measures, including read-only operations.
Here’s a video that explains how to connect your Zerodha account to AI assistants using Kite MCP:
Join our community by subscribing to our Weekly Newsletter to stay updated on the latest AI updates and technologies, including the tips and how-to guides. (Also, follow us on Instagram (@inner_detail) for more updates in your feed).
(For more such interesting informational, technology and innovation stuffs, keep reading The Inner Detail).







