Think inside your AI world.

ClickHouse through Unl

ClickHouse's MCP server will run the query and hand back the number. Whether that number is fit to sit on an executive dashboard is a freshness call you make separately - a query that resolves isn't the same as a pipeline that kept pace.

ClickHouse through Unl reads live row counts and query results against the freshness tolerance you have ratified, so 'can we trust today's number?' comes back pass-or-hold with the condition it hasn't met.

What ClickHouse holds

Run a query through ClickHouse's MCP server and it returns:

  • query results from any read-only SELECT against the cluster
  • row counts and aggregates computed live
  • the full list of databases on the cluster
  • table listings within a chosen database

What the naked read gives you

ClickHouse's MCP server runs the SELECT, lists the databases, lists the tables - and returns whatever numbers the query produces, whether that is a row count, a sum, or a percentile. It reports the cluster's data faithfully, with no view on whether today's number looks like a normal Tuesday or a broken pipeline.

The frame judges the data it is given; it does not verify the source’s accuracy.

What changes when ClickHouse is read measured

A BI lead at an events-ticketing company ratified a freshness rule in Unl: no revenue figure reaches the executive dashboard unless the day's event row count sits within five percent of the trailing seven-day average by ten each morning, because a stalled ingestion pipeline once let a flat Tuesday go unnoticed until quarter close.

'Can we trust today's revenue number?' returns, measured: not yet - the row count query shows 18,204 events logged for today against a seven-day average of 22,150 at the same hour, an eighteen percent shortfall past the five percent tolerance. ClickHouse supplied the row count; your ratified rule supplied the tolerance.

And back again

Once ingestion catches up, the recovered row count shows up the next time ClickHouse through Unl looks; the ratified freshness rule clears the number for the dashboard, revising the same page.

The answer comes back measured against what you already decided, and why.

The lane is live and open to this tool today: one box, paste anything. If it speaks MCP, Unl can reach it. Readings arrive unprompted, the data beside the criterion; Unl is a courier, not a warehouse, and keeps only your keys and the frame.

Questions people ask

What would make today's number trustworthy?

The row count catching up to within five percent of the seven-day average. The same ratified freshness rule reads the recovered count and clears the number for reporting.

How do I connect ClickHouse to Claude?

Run the official ClickHouse MCP server (ClickHouse/mcp-clickhouse), point it at your cluster with host and credentials, and register it with Claude. It runs in read-only mode by default, and Unl's reads never touch write mode.

Does Unl write to my cluster?

No. Unl calls run_select_query, list_databases, and list_tables - all read-only by default on the ClickHouse server. Any pipeline fix, you or your ingestion team apply it.

What this is

Think inside your AI world — you stay in command

Unlimitless (Unl to friends) holds what you've settled, reads what your tools are showing, and catches what's changed out in the world — and hands your AI whatever bears on the work, the moment it's needed, without you asking. The right thing, in front of the model, unprompted, with you in command of the call. So you keep moving toward what you set out to build, on top of everything you've already decided.

It plugs into Claude, Claude Code, ChatGPT and Cursor as an MCP connector. Quick to connect, in a couple of steps.

Unlimitless is open now to invited Alpha. Apply for the Beta waitlist to come in ahead of the full launch:

Alpha is invite-only · free at launch.