About
We build reviewable workflows.
Purpose
What we are building toward.
AI agents are moving from answering questions to doing work: reading data, calling tools, preparing transactions. The more they do, the more it matters that a person can see what is about to happen, stop what should not, and prove afterwards what did.
Cortiq is built around that idea. Every task becomes a plan you can read. Every tool has a fixed job and a label. Every action waits for a human decision. Every run can leave a small, permanent receipt on Base without putting its content on chain.
Today
What already works.
Reason, Flow, Console and Verify all run in your browser. Reading Base goes through a read-only relay on this site; nothing else about your runs reaches our server, and no account is needed.
The receipt contract is live on Base, so anchoring sends a real transaction from your own wallet. Hosted runs, outside integrations and $CORTIQ credits are planned and are labelled as planned wherever they appear.