Can Memtro change anything in ClickHouse?

Not yet. Today every ClickHouse tool reads, searches or lists. When the connector gains actions, each is marked, granted by name per team and used only when a person asks for exactly that, so nothing is created, sent, updated or deleted until you say so.

Who can use a ClickHouse connection?

A personal connection serves only your own requests. A shared connection serves the organisation: everyone, or only the teams an admin chooses. Admins can always add shared connections; other members need the permission.

Does Memtro store my ClickHouse data?

Credentials are encrypted at rest. Tool results are cached briefly so agent loops do not repeat identical calls, and the facts the model extracts are saved to your organisation's memory with provenance (system, record and link) so they can be checked and superseded.

How do I connect ClickHouse?

User + password (HTTP interface). Create a read-only user: CREATE USER memtro IDENTIFIED BY '…' SETTINGS readonly = 1; GRANT SELECT ON yourdb.* TO memtro;.

ClickHouse + Memtro

Ask questions of your ClickHouse data in plain English, answered with real SQL.

Memtro connects to ClickHouse Cloud or a self-hosted server over its HTTP interface with a read-only user. It browses databases, tables and columns, runs single SELECT statements with readonly=1, a 20-second execution cap and a 500-row limit, and can explain a query plan before running something heavy. The ClickHouse SQL dialect is described to the model so time-window and aggregate queries come out right.

Connect ClickHouse Start for free Reference

What you can do with ClickHouse in Memtro

Ask in plain English from chat, your editor, a scheduled job or your own agent. The model picks the right ClickHouse tool, reads what it needs, and answers with the source cited.

01

Product analytics on demand

A product manager asks how feature adoption changed after launch. Memtro inspects the events table, writes a ClickHouse query with toStartOfDay and uniq, runs it read-only and reports the daily numbers with the SQL it used.

02

Safe exploration of a big table

An analyst wants a breakdown over a 2-billion-row table. Memtro runs EXPLAIN estimate first to show how many parts and rows would be read, then runs the query with the execution cap so a slow scan is cut off rather than tying up the cluster.

Databases and tables with engine, row count and size
Columns with types, keys and sample rows
Read-only SQL with server-side readonly=1, execution and row caps
Query statistics: elapsed time, rows and bytes read
EXPLAIN plan, estimate, syntax or pipeline without running the query

Ask it like this

Questions people ask once ClickHouse is connected. In agent mode each one also checks the other systems that know about the same people, companies or identifiers.

›Which tables in the events database are the largest?
›How many page views per day in the last 14 days?
›What columns does analytics.sessions have?
›Top 10 countries by unique users this week
›Explain the plan for this query before I run it on the full table
›Average order value by hour yesterday

Where it fits

Teams that get the most from ClickHouse in Memtro, with the rest of their stack attached.

Better together

A record in ClickHouse is rarely the whole story. Connect the systems that hold the rest and the agent cross-references them by default, then remembers what it learned so the next question starts further along.

Also in Databases and warehouses

Using something else for the same job, or both? Each connector has its own tools; connect as many as you like.

How to connect

01Create a read-only user: CREATE USER memtro IDENTIFIED BY '…' SETTINGS readonly = 1; GRANT SELECT ON yourdb.* TO memtro;
02Grant SELECT on system.tables, system.columns and system.databases so Memtro can browse the schema
03In Memtro choose ClickHouse, enter the HTTPS endpoint (port 8443 on Cloud, 8123 self-hosted), user, password and default database
04Click Test and connect

Also works from

Chat in the dashboard, the OpenAI-compatible API, scheduled jobs, and any MCP client: Claude Code, OpenCode, Cursor, Claude Desktop. The ClickHouse tools are the same everywhere.

Set it up as an MCP server →Clients and agents →

Questions

Can Memtro change anything in ClickHouse?: Not yet. Today every ClickHouse tool reads, searches or lists. When the connector gains actions, each is marked, granted by name per team and used only when a person asks for exactly that, so nothing is created, sent, updated or deleted until you say so.
Who can use a ClickHouse connection?: A personal connection serves only your own requests. A shared connection serves the organisation: everyone, or only the teams an admin chooses. Admins can always add shared connections; other members need the permission.
Does Memtro store my ClickHouse data?: Credentials are encrypted at rest. Tool results are cached briefly so agent loops do not repeat identical calls, and the facts the model extracts are saved to your organisation's memory with provenance (system, record and link) so they can be checked and superseded.
How do I connect ClickHouse?: User + password (HTTP interface). Create a read-only user: CREATE USER memtro IDENTIFIED BY '…' SETTINGS readonly = 1; GRANT SELECT ON yourdb.* TO memtro;.

Connect ClickHouse in a few minutes

Seven-day free trial, then one monthly price per organisation. Connectors, jobs and memory are never metered.

Start for free Pricing