Concepts
Streams
Last updated: May 21, 2026
Streams are how Q (your friendly AI teammate) learns how to do part of your work.
Look back at your week. There are tasks you do over and over again. A weekly update every Friday. Proposals that go out to new customers. Performance reviews at the end of each quarter. Each one follows a pattern. You take similar steps, you consider similar things, and that's how you create work that feels like yours. Even if you've never written any of it down.
Anytime a proposal pops up, you do it a certain way. As you teach Q how you do proposals (or whatever work you're doing), Q writes all of that down in a Stream. Each Stream holds a playbook (the steps you follow), specs (the nitty gritty how to), and samples (examples to learn from).
Most of those files are simple Markdown. You can edit them yourself. You can take them to any other tool or chat. You can share them with coworkers. They're yours, just written down in a way that AI can use them to help you do your work.
From there, you assign Tasks to your Stream and Q does them for you. You can watch and manage this work. When Q gets things wrong, you give it feedback, and your Stream gets smarter.