MCP
A step-by-step guide to getting started with Nekt’s MCP.
Welcome to Nekt’s MCP!
Nekt’s MCP is the easiest way to interact with your data using natural language. When integrated with your agent, it provides a structured context and a set of tools, empowering it to reason, act, and respond effectively.
This guide will walk you through the setup process to get your MCP server up and running quickly.
Video Tutorial
For a visual walkthrough of the setup process, watch this video:
Prerequisites
You must have your cloud environment updated to version 2.0.19 or later.
AWS Costs
Your AWS costs will vary by region. For example, in Northern Virginia (us-east-1), the fixed cost is $34.00 per month.
Here is a list of the services used:
Step 1: Enable Your MCP Server in Cloud Settings
- Enable the MCP server by clicking “Enable MCP Server” button.
- Choose a custom domain you own and enter it here. This will be the endpoint for your MCP server.
Step 2: Verify Your MCP HTTPS Certificate with Your DNS Provider
Add the CNAME record from Nekt in your DNS provider’s dashboard to validate your HTTPS certificate. Please allow 5-10 minutes for this change to be verified.
Step 3: Add Your MCP URL to Your DNS Provider
While still in your DNS provider’s dashboard, add the provided CNAME record to point your domain to the new MCP server URL. Please allow an additional 5-10 minutes for verification.
Your MCP is now set up and ready to use.
Step 4: Create Tokens
You can now create multiple tokens to access your MCP server. For each token, you can associate any number of tables. Table scope is further mentioned in “Available Tools” section below.
Available Tools
The MCP provides three tools for your agent: ask_a_question, execute_sql_query, and save_assistant_response.
Tool: ask_a_question
You can use this tool to ask a natural language question about your data. The tool generates an Athena SQL query, which can then be executed using the execute_sql_query tool to retrieve the answer.
The scope of this tool is limited to the tables selected when you created the token.
Tool: execute_sql_query
Executes an Athena SQL query to retrieve data from your data lake.
Tool: save_assistant_response
This tool should be called at the end of a workflow after the agent has formulated its final response. It appends the response to the session’s history file, making it available for context in subsequent interactions.
Need Help?
If you encounter any issues during the setup or need assistance, feel free to reach out to our support team. We are here to help you get started.