Think inside your AI world.

Looker through Unl

Looker runs the Explore and returns the data. Whether it's answering the question that was actually asked is your rule - and 'the numbers came back' isn't the same as 'the numbers use the definition we agreed on.'

Looker through Unl reads Explore, Look and dashboard results against the metric definition you have ratified, so 'does this answer the question' comes back in-or-out with the condition it hasn't met.

What Looker holds

What Looker through Unl actually reads, direct from the Looker MCP toolbox:

  • LookML models, Explores, dimensions, measures, filters and parameters
  • Live query execution against an Explore, with data returned
  • Saved Looks matched by title or description, and their results when run
  • Saved dashboards matched by title or description
  • The generated SQL and a direct link back to the query in Looker

What the naked read gives you

Ask Looker through Unl to run the 'active customers' Explore and it returns real numbers, computed live against the semantic model - the dimensions, the filters applied, the SQL Looker generated to get there. It is an honest read of what that Explore, defined exactly as it is today, produces. What no Explore settles on its own is whether 'active customers' as defined there is the same 'active' the CFO meant when they asked the question - Looker runs whichever definition is wired into the model, with equal confidence, whether or not it's the right one for this conversation.

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

What changes when Looker is read measured

Say you're a revenue operations analyst who fields exec data requests, and you've ratified that an 'active customer' figure only satisfies a board question if it uses the Explore tagged finance-approved and excludes the trial-account filter - any other active-customer Explore gets flagged as a different metric, not a wrong one.

'Does this Look answer the CFO's active-customer question?' returns, measured: no, because the Look runs against the growth-team Explore, which includes trial accounts your rule excludes from the finance definition. Looker supplied the data and the Explore it came from; your ratified rule supplied which Explore counts as the answer.

And back again

Re-run the question against the finance-approved Explore, then log the result to Unl against your definition rule - the next person who asks the same thing inherits which Explore actually answered the CFO, not just a number that happened to look similar.

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

Why did Looker through Unl say this Look doesn't answer the question?

Because it runs against an Explore your ratified rule doesn't recognise as the finance definition, even though the query executes and returns numbers. Point it at the approved Explore, or change the rule, and the verdict follows.

How do I connect Looker to Claude?

Enable the Looker-managed MCP server from your Looker instance's admin panel, or run the open-source MCP Toolbox for Databases against your instance. Either way, Unl reads Explores, Looks and dashboards through that same connection.

Does connecting Looker let Unl change my Explores or LookML?

No. Unl runs read queries, retrieves saved Looks and dashboards, and reads the LookML model - it doesn't edit the semantic layer or your saved content.

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.