Changelog
What has changed in Unl, newest first. Each change links to the page where you can see it.
Every change merged into Unl, counted by day since May, is drawn on our GitHub page. To follow the build, star unl on GitHub.
A clearer way around the site
The top of every page now has Product, Solutions and Resources menus, then Pricing and Docs. The footer is grouped the way you would look for things: product, who it is for, resources, company and legal.
Go to the homepageFollow one decision through the work
Unl Loops follow one real conversation from the judgement made in it, through the instruction it became and the agent's action, to the changed page and the evidence that came back.
Watch a LoopEvery way to add Unl, on one screen
Connect now walks you through adding Unl to each AI you use, step by step, in one place.
Connect your AIFor builders after the build
A page for people whose product is real and whose next job is the business: researching customers, testing a message and choosing the next change, all from the same decisions.
Read itUnl in Slack
When your agents need a new decision, the conversation can happen in Slack. Answer in words, and the reason is kept for the next agent.
See how it worksSee Unl at work
A page that replays how we build Unl with Unl: each AI at work, the decisions Unl handed it, and why.
See Unl at workBuilt for orchestration
How Unl carries your why into every agent an orchestrator starts, shown on the fleet we run ourselves.
Read itSupport from a person
One address, read by a person, and the pages that answer the common questions in one place.
Get supportSecurity, in plain words
Our security commitments, rewritten so each one is short, plain and true today.
Read our commitmentsCodex on the install row
Add Unl to Codex from the same row as the other AIs, now arranged as chat surfaces, coding surfaces and any model.
Add Unl to your AIHow Unl compares
Custom instructions and rules files each do part of the job. This page sets out where each one stops and what Unl adds.
CompareYour tools, read against what you decided
Connect the tools you already use once in your portal, then ask in whichever AI you work with. Unl brings the answer back measured against the decisions you have kept.
See your tools