Zoho DataPrep MCP Server
Overview
What is MCP?
Model Context Protocol (MCP) is an open protocol that establishes the standardized manner in which applications can interact with Large Language Models (LLMs). An MCP allows you to carry out your desired workflow on your preferred LLMs by providing the required instructions in the right context to an AI model.
With an MCP in place you can instruct popular AI agents or even custom AI agents to interact with native or third-party applications and complete complex workflows with little to no human intervention.
Zoho DataPrep Model Context Protocol (MCP) server
Zoho DataPrep Model Context Protocol (MCP) server enables teams to prepare, transform, import, and export data using conversational AI. It allows you to interact with Zoho DataPrep through natural language from AI interfaces like Claude without manually navigating the UI or writing API calls. You can create pipelines, manage stages, execute pipelines, and fetch results simply by describing what you want. Zoho DataPrep MCP is built on the Zoho MCP framework, so you must access Zoho MCP and configure a server to enable DataPrep as an MCP tool. Click here to learn more about Zoho MCP
Steps to add Zoho DataPrep as a tool in Claude using MCP
Access Zoho MCP server
1. Go to https://www.zoho.com/mcp/ and request playground access.
2. Once, the request is approved, go to https://mcp.zoho.com/.
3. Click on Create MCP server option and give a name for your server and click Create.
4. A server will be created. Click on Tools option in the side pane add Zoho DataPrep to your server.
5. Select the tools as required and click Add now.
6. Next we need to add Zoho DataPrep MCP to the LLM. To do this, click on Connect in the side pane and copy the MCP URL value
Link Claude with Zoho DataPrep
Link Claude with Zoho DataPrep MCP, allowing it to handle real-world workflows through a secure, straightforward setup—all within your conversation window.
Setting Up Zoho DataPrep MCP in Claude
1. Go to Claude and open the Settings.
2. Click on the Connectors option.
3. Select Add custom connector.
4. Enter the name and paste the MCP URL that was copied in step 6 and click Add
Access Zoho DataPrep MCP Tools in Claude
1. Open a new chat.
2. Choose Search and tools.
3. Pick the Zoho DataPrep MCP connection.
4. Select the tools you’d like to use in the chat.
5. Ask Claude to perform actions using them.
Zoho DataPrep MCP tools
The below table lists the various Zoho DataPrep MCP tools that we support
Tool Name | Description |
Add New Connection | Creates a new connection. |
Add Pipeline to Workspace | Creates pipelines under a specific project. |
Create New Pipeline | Creates new pipelines. |
Create Workspace | Creates new projects. |
Delete Workspace | Deletes existing projects. |
Import Pipeline Catalog | Initiates import/refresh of a pipeline’s catalog and metadata. Returns job ID and updated task details. |
getPipelineImportCatalog | Retrieves extract/import task details required for pipeline runs and schedules. |
Get all Workspaces | Retrieves all projects. |
getSubscription | Retrieves subscription details, usage metrics, plan limits, addons, and expiry information. |
Get Stage details | Retrieves dataset metadata, structure, sample data, quality metrics, and statistics. |
Re-authenticate Connection | Re-authenticates an existing connection. |
Refresh Stage Metadata | Updates and refreshes the dataset metadata and quality metrics. |
Remove Pipeline Schedule | Removes a pipeline’s schedule and stops future automated runs. |
Retry Pipeline Run | Retries a failed pipeline run by creating a new run with the same configuration. |
Run Pipeline | Executes a pipeline. Supports MANUAL, RELOAD, BACKFILL, and MANUAL_WITH_DATA run types. |
Schedule Pipeline | Creates an automated schedule for pipeline execution. |
See Pipeline Jobs for this Portal | Retrieves a paginated list of all pipeline job runs with filters for status, type, date, etc. |
Update Connection | Updates an existing connection. |
Bulk Update Data Stage | Updates multiple datasets. |
Update Data Stage in a Workspace
| Updates datasets under a specific project. |
Update Pipeline | Updates an existing pipeline. |
Update Pipelines in a Workspace
| Updates pipelines under a specific project. |
Update Pipeline Schedule | Updates an existing pipeline schedule configuration. |
Update Workspace | Updates project details. |
View All Pipelines | Retrieves all pipelines. |
View All Pipelines in a Workspace | Retrieves pipelines within a specific project. |
View All Stages | Retrieves all datasets. |
View All Stages in a Workspace | Retrieves datasets within a specific project. |
View Connection Details | Retrieves details of a specific connection. |
View Default Portal Details | Retrieves default portal details |
Use cases
Here are a few example prompts you can use in Claude.
1. Failed Pipeline Debugging & Retry
Use case: User wants to understand why yesterday’s job failed and fix it.
Sample Prompt: My sales pipeline failed last night. Show me the exact error and retry the failed run.
Tools involved:
- Gell all pipeline jobs for portal
- Get pipeline run info
- Get pipeline info
- Retry pipeline run
2. Schedule Management for Production Pipelines
Use case: User wants to automate and control pipeline execution.
Sample Prompt: Schedule my Sales pipeline to run every day at 2 AM, and if a schedule already exists, update it.
Tools involved:
- Get pipeline schedule
- Schedule pipeline
- Update pipeline schedule
3. Connection Failure & Re-authentication Flow
Use case: Production pipeline failed due to expired credentials.
Sample Prompt: My Sales pipeline is failing due to authentication issues. Re-authenticate the connection and rerun the pipeline.
Tools involved:
- Get connection
- ConnectionReAuth
- Run pipeline