StarRocks MCP. Manage OLAP data, audit clusters via AI query.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
StarRocks MCP Server connects your AI agent directly to a high-performance OLAP database. It lets you query data, check cluster health, and manage schemas using natural language prompts—no SQL terminal needed.
Your agent can list databases, retrieve full table structures via `get_table_schema`, monitor node status with `list_nodes`, or run complex analytical queries using `execute_query`.
This is for advanced database operations.
What your AI agents can do
Execute query
Runs any arbitrary SQL query against your StarRocks data set.
Get cluster info
Retrieves general status and metadata for the frontend nodes of the cluster.
Get storage usage
Checks the total disk usage and data size statistics across all cluster components.
List all databases, then retrieve the specific schema or list tables within any given database.
Execute custom SQL queries on demand to pull real-time metrics and insights from your data warehouse.
Check the status and metadata of all cluster nodes (FE and BE) or get overall system info via get_cluster_info.
List active data load jobs (list_jobs) and review materialized view definitions to track data movement.
Get total disk usage or specific storage statistics across the entire distributed cluster using get_storage_usage.
Ask AI about this MCP
Supported MCP Clients
OAuth 2.0 Compatible Gemini Waiting for input…
StarRocks MCP Server: 10 Tools for Database Operations
These tools give your agent the ability to list databases, check node status, retrieve schemas, and execute any arbitrary query needed to manage your OLAP infrastructure.
Make your AI actually useful.
Add this MCP to Claude, Cursor, or Windsurf and your AI stops guessing. It gets real tools to look things up, take action, and handle the stuff you keep doing by hand.
Start using StarRocks on Vinkius019d8484execute query
Runs any arbitrary SQL query against your StarRocks data set.
019d8484get cluster info
Retrieves general status and metadata for the frontend nodes of the cluster.
019d8484get storage usage
Checks the total disk usage and data size statistics across all cluster components.
019d8484get table schema
Pulls the full structure, including columns and types, for a specified table.
019d8484list databases
Returns a list of all databases available in the StarRocks cluster.
019d8484list jobs
Shows a list of active or completed data load jobs running through your pipelines.
019d8484list mvs
Lists all defined materialized views used for accelerating analytics queries.
019d8484list nodes
Provides detailed status information for every backend node in the cluster.
019d8484list tables
Lists all tables that exist inside a specified database.
019d8484list views
Retrieves a list of views defined within a specific StarRocks database.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with StarRocks, then connect any of our 4,900+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,900+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by StarRocks. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 10 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Debugging complex data pipelines used to take an hour of clicks and CLI work.
Today, if your data pipeline fails, you open the StarRocks Manager GUI. You click on 'Jobs,' then drill down into a specific job ID. Then you might have to manually check logs or run separate queries just to see which table failed and why. It's a multi-step process involving five different screens.
With this MCP Server, you tell your agent: 'Check the status of yesterday's data load.' The agent runs `list_jobs`, checks for failures, and uses `get_storage_usage` to confirm if the expected amount of data arrived—all in one chat exchange. You get the answer fast.
StarRocks MCP Server: 10 Tools for Database & Cluster Ops
You don't have to remember dozens of SQL commands or navigate deep menus just to check basic health. You can ask the agent, 'Are all my backend nodes online?' and it runs `list_nodes` instantly, giving you a clean pass/fail report.
The difference is control. Instead of being limited by what the GUI designer thought was necessary, your AI client gives you direct, conversational access to every core operation: schema checks, query execution, node monitoring, and more.
What you can do with this MCP connector
Forget wrestling with a terminal or writing boilerplate SQL just to check your data warehouse health. This server hooks up your agent straight to StarRocks, letting you run complex database operations using natural language—no command line needed. Your AI client acts like an expert DBA sitting right next to you.
Discovering the Data Structure
You can start by listing every database available in the cluster with list_databases. Once you've got a DB name, you pull a list of all tables inside it using list_tables, or check out defined views via list_views. To know exactly what columns you’re working with, run get_table_schema on any table to pull its full structure and data types.
Running Queries & Analyzing Insights
When you need answers, just use execute_query and run any SQL query against your entire StarRocks data set instantly. If you're optimizing performance, check out all defined materialized views using list_mvs. You can also track the lifeblood of your platform by viewing a list of active or completed data load jobs through list_jobs.
Monitoring Cluster Health
To make sure the platform isn't coughing up smoke, you can check overall system status and general metadata using get_cluster_info. For a detailed look at the actual backend components, list_nodes provides specific status information for every node in the cluster. If you gotta know how much space your data is eating up, run get_storage_usage to get disk usage and total data size statistics across all cluster parts.
019d8484-b82a-7173-9daa-5b0d62d872c7 
How StarRocks MCP Works
- 1 First, subscribe to the StarRocks server and provide your FE Host, Port, User, and Password credentials.
- 2 Next, talk to your AI client (Claude, Cursor, etc.) using natural language. Ask it to perform a task, like 'Show me all databases.'
- 3 Finally, the agent uses the appropriate tool (e.g.,
list_databases), gets the output, and presents you with actionable data or results.
The bottom line is that your AI client translates complex database actions into conversation and executes them for you.
Who Is StarRocks MCP For?
Data Engineers who hate manual schema audits. Analytics Leads who need to validate data access without writing boilerplate SQL. Infrastructure SREs tired of checking multiple dashboards just to confirm node health.
Automating repetitive tasks like auditing schemas or monitoring the status of ETL ingestion pipelines.
Quickly verifying data access and materializing views by asking natural language questions about data structures.
Monitoring the health of the cluster nodes and checking overall storage usage without logging into a terminal.
What Changes When You Connect
- Audit your entire schema instantly. Instead of running
SHOW DATABASES;and then manually checking each one, just ask the agent to list databases usinglist_databases. It handles the whole flow. - Check cluster health without logging into a dashboard. The
get_cluster_infotool gives you immediate status on your FE nodes, letting you confirm everything is running fine in seconds. - Know exactly what data exists before querying it. Use
list_tablesand thenget_table_schemato validate the structure of any table before writing a single line of SQL. - Track ETL pipelines reliably. The agent can use
list_jobsto show if your recent data loads succeeded or failed, saving you from manually checking job logs. - Get an immediate storage breakdown. Running
get_storage_usagelets you see disk space consumption across the entire distributed cluster in one single query.
Real-World Use Cases
Schema Validation After a Data Source Change
A data engineer needs to confirm if a new source table, raw_user_data, was added to the 'staging' database. Instead of running 10 separate checks, they prompt their agent: 'What is the schema for raw_user_data?' The agent calls get_table_schema and instantly delivers the full column list, confirming the data structure before they write any transformation code.
Investigating a Slow Query
An analytics lead notices that reports are slow. They first ask the agent to check the cluster status using get_cluster_info and then run list_nodes. If all nodes look healthy, they use execute_query with specific monitoring SQL to pinpoint if a particular query is hitting an unindexed column.
Auditing Materialized Views
An SRE needs to know which materialized views exist and how often they run. They prompt the agent: 'List all MVs in the sales database.' The agent uses list_mvs and then confirms the definition details, ensuring no critical data view was missed during maintenance.
Pre-Deployment Data Check
A developer is about to deploy a new dataset. They first ask the agent to list_databases to confirm the target environment is correct, then use get_storage_usage to estimate if the incoming data will exceed allocated disk space before running any inserts.
The Tradeoffs
Running a Query Blindly
The user runs an arbitrary query using execute_query and gets an error message saying 'Column X does not exist.' They then try to manually fix the SQL statement without knowing why the column was missing.
→
Before running any query, always check the structure. Use the agent first to run get_table_schema on your target table. This confirms the exact column names and data types you should use in your SQL.
Confusing Nodes with Databases
The user thinks running 'list nodes' will show them all their available databases, confusing infrastructure status with logical structure.
→
Remember the difference. To see data containers, use list_databases. If you need to check if your underlying hardware/FE services are up, use list_nodes.
Guessing the right table name
The user tries to query a table called 'sales' but StarRocks contains two tables: 'sales_us' and 'sales_eu'. They guess wrong, and the query fails.
→
Always start by listing what's there. Use list_databases first, then use list_tables within that database to see every available table name.
When It Fits, When It Doesn't
Use this server if your job involves deep data auditing, monitoring complex cluster health, or running ad-hoc queries against a high-volume OLAP system. You need tools like get_table_schema, list_nodes, and execute_query to manage the backend infrastructure itself.
Don't use it if you just need simple CRUD operations (Create/Read/Update). If your goal is simply generating a report from clean, pre-structured data, a dedicated BI tool might be faster. Only rely on this MCP Server when the complexity of the data structure or the monitoring task requires interaction with the database engine's metadata itself.
Common Questions About StarRocks MCP
How do I find out all available databases using list_databases? +
Just ask the agent to 'list all databases.' The tool list_databases runs automatically and gives you a clean list of every database name in your cluster, so you know what scope you're working with.
What is the difference between list_nodes and get_cluster_info? +
They provide different levels of detail. get_cluster_info gives you general system status for the frontend nodes, while list_nodes provides deep, specific metadata about every individual backend node.
Can I check a table's schema without listing databases first? +
No. The agent must establish context. You should start by using list_databases to narrow down the scope, and then use list_tables within that database before asking for the specific structure with get_table_schema.
How does the execute_query tool work? +
The execute_query tool runs standard SQL queries. You simply tell your agent what data you need, and it formats and executes the query for you, returning the results directly in the chat.
How do I use list_jobs to monitor the status of data ingestion pipelines? +
It lists all current and historical data load jobs. You can check which jobs are running, their completion status, or review past performance records for your analytical pipelines.
How do I use get_storage_usage to audit disk space across my StarRocks cluster? +
This tool retrieves detailed statistics on data storage consumption. It shows you the total disk usage and size metrics, letting you spot which part of your distributed cluster is running low on capacity.
After listing a database, how do I use list_tables to get all table names? +
It outputs a simple list of every table housed within the selected database. This helps you quickly enumerate available tables before needing to run schema checks or query data.
What is the purpose of list_mvs, and how do I check for materialized views? +
list_mvs lists all existing materialized views in your cluster. You use these MVs to pre-calculate common metrics, which significantly speeds up complex queries that rely on aggregated data.
How do I find my StarRocks FE Host and Port? +
The FE Host is the IP address or domain of your StarRocks Frontend node. The default HTTP port is 8030. You can verify these in your deployment configuration or by contacting your administrator.
Can I execute DELETE or UPDATE queries through this server? +
Yes. Use the execute_query tool. As long as your authenticated user has the necessary permissions in StarRocks, you can execute any valid SQL statement, including DDL and DML operations.
Is it possible to monitor cluster health through the agent? +
Yes! Use the get_cluster_info and list_nodes tools to retrieve real-time status for all Frontend and Backend nodes, helping you identify offline or overloaded nodes instantly.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.