Zetaris MCP Server
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Zetaris MCP Server list the tables in the montgomery database and describe the census table"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Zetaris for Cursor
Connect Cursor's agent to your Zetaris instance, so it gets data through Zetaris instead of going around it.
After clicking, replace the sign-in values in Cursor's MCP settings with your own. The button runs the server with uv (uvx).
What you get
zetaris_mcp.py: a small MCP server (Python standard library only). It gives the agent five tools:list_databases: see what data Zetaris already has.list_tables: see the tables in one database, plus any ready-made views you describe.describe_table: see a table's columns.run_query: run read-only SQL (SELECT,WITH,SHOW,DESCRIBE,EXPLAIN).register_source: add a new source (CREATE LIGHTNING ...) or cache a table (CACHE TABLE).
.cursor/rules/zetaris.mdc: a project rule Cursor reads before every task. It tells the agent to get data through Zetaris, register a missing source instead of downloading it, join sources inside one Zetaris query, and show the SQL behind every figure.examples/montgomery/: the rule and view list used for Montgomery, Alabama (City of Montgomery, US Census, CDC and Ookla data), as an example of describing your own ready-made views.
Related MCP server: mcp-odbc
Set up by hand (no uv needed)
Copy
zetaris_mcp.pyand the.cursor/folder into your project.Create
.env.localnext to them (it is in.gitignore):ZETARIS_BASE_URL=http://localhost:3000 ZETARIS_ORG_ID=1 ZETARIS_USERNAME=you@example.com ZETARIS_PASSWORD=...Or set
ZETARIS_API_KEYinstead of the username and password.Open the project in Cursor. In Settings → MCP, turn on
zetaris. With the Cursor CLI:cursor-agent mcp enable zetaris.Ask a question in plain English. The agent finds the data in Zetaris, writes the SQL and shows it with the answer.
Each tool call is written as one JSON line to zetaris-mcp.log, so you can see exactly what the agent ran in Zetaris.
Settings
Variable | Default | Meaning |
|
| Your Zetaris UI address |
| none | Numeric organisation ID |
| none | How to sign in |
| none | File to read these from (the bundled |
|
| Seconds each call may take before the agent is told to make the query lighter |
| none | Spark status address (for example |
|
| Longest wait, in seconds, before a query that ran past the time limit is assumed finished |
| none | JSON file describing ready-made views, shown by |
|
| Most rows returned per query |
| none | Where tool calls are logged |
Safety
run_queryrefuses anything that is not read-only.register_sourceonly acceptsCREATE LIGHTNINGandCACHE TABLE. Neither tool can drop or change existing data.Each call has a time limit, so the agent gets a clear answer instead of hanging.
With
ZETARIS_SPARK_UIset, a query that ran past the limit blocks new queries until Spark shows it has finished. An agent cannot pile retries onto a busy engine and slow down other users of the same Zetaris.
Licence
MIT
Available Tools
5 toolsdescribe_tableB
Show the columns and types of a Zetaris table, given as DATABASE.table.
Inputs: table (required).
list_databasesA
List the databases registered in Zetaris. Start here to see what data is already available.
Inputs: no named parameters.
list_tablesC
List the tables in one Zetaris database.
Inputs: database (required).
register_sourceB
Register a new data source in Zetaris with a CREATE LIGHTNING ... statement (REST, FILESTORE or DATABASE), or cache a table with CACHE TABLE.
Inputs: sql (required).
run_queryA
Run a read-only SQL query in Zetaris (SELECT, WITH, SHOW, DESCRIBE, EXPLAIN). Zetaris SQL is Spark SQL; tables are DATABASE.table and one query can join tables from different sources. Returns up to 200 rows, the time taken and the exact SQL run.
Inputs: sql (required).
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 5 tools
The tools mostly target distinct operations: listing databases, listing tables, describing a table, querying, and registering a source. There is some overlap between describe_table and run_query because run_query also accepts DESCRIBE, but the descriptions make the intended use cases reasonably clear.
All tool names use a consistent snake_case verb_noun pattern: list_databases, list_tables, describe_table, run_query, register_source. The convention is predictable and easy to infer.
Five tools is well-scoped for a Zetaris data exploration and query server. Each tool covers a distinct step in the discovery-to-query workflow, with no obviously redundant tools.
The server covers the core lifecycle of discovering databases, inspecting tables, querying data, and registering sources. Minor gaps remain around removing or updating registered sources and inspecting source metadata, but agents can handle most core workflows.
Maintenance
Related MCP Connectors
Text-to-SQL MCP server: read-only queries on PostgreSQL, MySQL and SQL Server with your real schema
An MCP server for search and retrieval over a local LanceDB database
Remote data science agents for Snowflake, Databricks & BigQuery in Claude/Cursor via MCP
Hosted SQL Server MCP server for Claude, ChatGPT and Cursor, with each action switched per account.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceA read-only MCP server for MS SQL Server, written in Python. Lets AI agents query your database safely — no writes, no stored procedures, no shell access.1-
- AlicenseNot gradedqualityDmaintenanceA Python MCP server that connects to any ODBC data source and exposes schema discovery + query tools. Read-only by default, write access available per-connection.2MIT
- AlicenseNot gradedqualityDmaintenanceRead-only MCP server that lets coding AI agents inspect Oracle Database schema through live metadata.ISC
- FlicenseNot gradedqualityCmaintenanceMinimal MCP server for read-only access to ClickHouse, enabling AI agents to explore schemas and execute SELECT queries safely.-