ClickHouse MCP Server
Connect to the ClickHouse MCP server to run SQL, manage services, and build dashboards and alerts using AI agents on Gumloop, Claude, or Cursor.
Try ClickHouse now
Type what you want done. Sign in and run it live with an AI agent.
|
ClickHouse
Installation
Set up the ClickHouse MCP server in Gumloop
Do this once to provision your hosted server URL.
Create a Gumloop account
To use this MCP, you need a Gumloop account. If you don't have one yet, you can create one for free.
Add and authorize the ClickHouse server
In Gumloop, open Connectors and add ClickHouse. Paste your ClickHouse API key when prompted. The credential is stored securely in Gumloop.
Then use it in your client
Use in GumloopUse ClickHouse in an agent
Once ClickHouse is set up, just open any Gumloop agent, add ClickHouse as a connector, and start chatting with the agent.
Tools (17)
Execute Query
Execute a SQL query against a ClickHouse Cloud service.
List Organizations
List ClickHouse Cloud organizations, or fetch one by id.
List Services
List services in a ClickHouse Cloud organization, or fetch one by id.
List Service Backups
List backups for a service (most recent first), or fetch one by id.
Get Service Backup Configuration
Get the backup schedule and retention configuration for a service.
List Clickpipes
List ClickPipes configured on a service, or fetch one by id.
Get Organization Cost
Retrieve daily billing and usage cost data for a ClickHouse Cloud organization (max 31-day window).
Update Service State
Start or stop a ClickHouse Cloud service.
List Dashboards
List ClickStack dashboards configured on a service, or fetch one by id.
Create Dashboard
Create a ClickStack dashboard.
Update Dashboard
Update a ClickStack dashboard.
Delete Dashboard
Delete a ClickStack dashboard.
What is ClickHouse MCP?
The ClickHouse MCP server gives AI agents access to your ClickHouse Cloud. That means agents can run arbitrary SQL against your services, list organizations and services, start and stop services on demand, inspect backups and streaming ClickPipes, pull daily billing and cost data, and create, update, and delete ClickStack dashboards and alerts. It works across your ClickHouse Cloud account, from the database layer up to the ClickStack observability plane.
If your team spends time juggling connection strings, writing one-off scripts against the ClickHouse Cloud API, or clicking through the console to check costs, restart a service, or tweak an alert, an AI agent can take over much of that busywork. Describe what you need, and your AI agent will handle the query, the service action, or the dashboard change for you.
MCP stands for Model Context Protocol. It’s an open standard that gives AI agents a way to connect to external tools and services. Instead of managing environment variables, wiring up HTTP Basic auth against two separate endpoints, and parsing query results yourself, you add your ClickHouse Cloud API key and database credentials once. After that, you can query and operate your ClickHouse Cloud just by chatting with your AI agent.
Related MCP servers
Google BigQuery MCP Server
Query and manage BigQuery datasets, tables, and execute SQL
Snowflake MCP Server
Interact with Snowflake data warehouse for executing SQL queries, managing tables/databases, and accessing data
PostgreSQL MCP Server
Query and manage PostgreSQL databases with schema exploration, SQL execution, and query analysis
Databricks MCP Server
Manage Databricks clusters, jobs, and job runs
Looker MCP Server
Interact with Looker Enterprise dashboards and analytics content
Tableau MCP Server
Interact with Tableau to access views, workbooks, datasources, and Pulse metrics
What you can do with ClickHouse MCP on Gumloop
Run any SQL as a database copilot
Execute arbitrary SQL against a ClickHouse service over its HTTPS endpoint: analytical SELECTs, INSERTs, and DDL like CREATE, ALTER, or DROP. Your AI agent can answer data questions, run schema checks with SHOW TABLES and DESCRIBE, and apply changes, with configurable output formats and timeouts.
Return results in the format you need
Choose the output format for each query (JSON, JSONCompact, CSV, and other ClickHouse formats). Results come back ready for an AI agent to summarize, chart, or pass straight into the next step of an automation.
List organizations and services
Pull every ClickHouse Cloud organization and the services inside it, or fetch a single one by ID. Your agent can report on your fleet without you opening the console.
Start and stop services for cost control
Send a start or stop command to any ClickHouse Cloud service. An AI agent can spin services down outside business hours and bring them back when work resumes, trimming idle spend.
Build and maintain ClickStack dashboards
Create dashboards with visualization tiles, update names, descriptions, tiles, filters, and tags, and delete dashboards you no longer need. This is full CRUD, so an agent can stand up observability views as code.
Set up and tune alerts
Create threshold-based alerts tied to a dashboard tile or a saved search, set above or below thresholds and check intervals, route notifications to a webhook or Slack, and enable, disable, update, or delete alerts. Alerts support full CRUD as well: create, update, and delete, plus list.
Track cost and billing
Pull daily billing and usage cost data for an organization over a date range (up to a 31-day window). Your AI agent can flag spend spikes and power cost-optimization recommendations.
Inspect backups, ClickPipes, and sources
List a service’s backups (most recent first) and read backup schedule and retention settings, list ClickPipes streaming ingestion connectors, and enumerate the ClickStack data sources available to attach dashboards and alerts. These are read-only views for monitoring and setup.
How to connect the Gumloop ClickHouse MCP Server
- 1
Create a free Gumloop account
Sign up at gumloop.com. No credit card required.
- 2
Add the ClickHouse MCP server
Copy your MCP server URL from Gumloop and add it to your preferred client (Claude, Cursor, or Gumloop workflows). You'll authorize on first use.
- 3
Start using ClickHouse in your AI workflows
That's it. Your AI agent can now run SQL, manage services, track cost, and build dashboards and alerts across your ClickHouse Cloud. Use it inside a Gumloop automation, in Claude Desktop, or in Cursor.
ClickHouse MCP use cases
Ad-hoc analytics for data analysts
Ask a question in plain language and let a Gumloop agent translate it into ClickHouse SQL, run execute_query against the right service, return the result as CSV or JSON, and post a summary to Slack or write it into Google Sheets. Analysts get answers from billions of rows without writing the query or exporting anything by hand.
Cost control for platform operations
A Gumloop agent can pull daily cost data for an organization, compare it against the prior period, and stop non-production services that are sitting idle, then start them again on a schedule. Platform teams keep ClickHouse Cloud spend in check without babysitting the console.
Observability-as-code for SRE and platform teams
List the ClickStack data sources on a service, create dashboards with the tiles you care about, and attach threshold alerts that page a Slack channel or webhook. When requirements change, an AI agent updates the tiles and alert thresholds, so your monitoring stays in sync with your infrastructure.
Incident response for on-call engineers
When something breaks, an agent can run diagnostic SQL against the affected service, check recent backups and their retention configuration, inspect ClickPipes ingestion connectors for a stalled stream, and disable noisy alerts while you work on incidents. Everything happens from one chat instead of a dozen console tabs.
Cross-platform data reporting for analytics teams
Connect ClickHouse with Slack, Gmail, Google Sheets, and other MCP integrations in a single agent. Run a scheduled query, format the results, drop them into a shared sheet, and send a digest email, so stakeholders get fresh ClickHouse numbers without anyone running the report manually.
Why use Gumloop for ClickHouse MCP
Your ClickHouse credentials, stored securely
Add your ClickHouse Cloud API key and database credentials once, stored securely. No env vars, connection-string juggling, or coding against the ClickHouse API necessary. Gumloop keeps your API key ID, secret, and database password safe so you don’t have to manage them in scripts or config files.
Works with multiple MCP clients
Use the ClickHouse MCP server endpoint in Claude Desktop, Cursor, or directly inside Gumloop agents. Same server URL, works with any MCP client.
Chain ClickHouse with 100+ other integrations
Combine ClickHouse with Slack, Gmail, Google Sheets, and other MCP integrations in a single AI agent. An agent can run a query in ClickHouse, process the results with an LLM, and write them to a sheet or send an alert anywhere else.
Enterprise-grade and scalable
Built for teams, with role-based permissions and dedicated support for Pro users. Whether you run one service or a large fleet, your AI agent operates within your existing ClickHouse access. For details on Gumloop’s security practices, see trust.gumloop.com.
Pricing includes a free plan
You can test the ClickHouse MCP integration on Gumloop’s free tier before committing. Paid plans start at $37/month.
Frequently asked questions
Ship ClickHouse agents in minutes
Connect any AI agent to 100+ MCP servers, zero setup.