Think inside your AI world.

Iterable through Unl

Iterable's MCP will hand back an experiment's metrics the second you ask. Unl asks first whether that lead clears the bar you ratified for calling a test won.

Iterable through Unl reads experiment and campaign metrics against the win threshold you have ratified, so 'did the test actually win?' comes back won-or-inconclusive with the condition it hasn't met.

What Iterable holds

What Iterable's MCP genuinely hands back, read-only by default:

  • Experiment metrics - variant performance for A/B testing analysis, via get_experiment_metrics.
  • Experiment listings - which tests are running and against which campaign, via list_experiments.
  • Campaign metrics - performance for a chosen date range, via get_campaign_metrics.
  • Variant content - what each experiment arm actually sent, via get_experiment_variants.
  • Template detail - the email, push, or SMS template behind a send, via get_email_template and related tools.

What the naked read gives you

Ask Iterable's MCP how an experiment is running and it tells you plainly - variant B is ahead by six points, current as of this morning. The server is not withholding anything or making the number hard to reach; by default it ships read-only, no PII, no writes, no sends, so the read itself is trustworthy. Nowhere in that read does the 5-point margin appear, the mark your team fixed between a real result and a variant that merely happened to be ahead when someone last checked.

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

What changes when Iterable is read measured

A lifecycle marketer ratified the rule in Unl: a test variant has to beat control by at least 5 percentage points and hold that lead across the full scheduled window before it becomes the new default, because rolling out a lead that evaporates by week two costs more in re-testing than waiting would have.

'Did the test actually win?' returns, measured: inconclusive - variant B is ahead by 6 points today, but the window has three days left to run and the lead has swung as low as 2 points mid-test. Iterable supplied the experiment metrics; your ratified rule supplied the margin-and-duration bar that keeps a mid-test lead from being called early.

And back again

The inconclusive call travels with the experiment rather than dissolving once the chat ends - Unl holds the 5-point bar and the current lead against this test, so when the window closes and the same question comes up again, whoever asks it finds the won-or-inconclusive history already there instead of everyone reconstructing where the lead stood three days earlier.

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

Would the verdict change once the test window closes?

Yes. The 5-point margin and full-window requirement stay fixed; what changes is the metric Iterable reports as the window runs out. If variant B still holds a 5-point-plus lead at the close, the same question returns won, not inconclusive.

How do I connect Iterable to Claude?

Iterable ships an open-source MCP server on GitHub; run its setup wizard, which defaults to read-only mode, and add the resulting connection as a custom connector in Claude, and Unl reads through that same server.

Does Unl launch campaigns or change experiments in Iterable?

Unl reads through Iterable, and can write back on your explicit gesture — it never acts as a side effect of a read.

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.