One year of AI Context Flow: from browser extension to open context layer
It’s been one year since our first launch, and that feels like a good moment to look back.
A year ago, AI Context Flow was a browser extension.
Today, it’s much closer to our original vision: an open context layer for people, teams and AI agents.
We’ve built it in public, one user conversation at a time. Here’s how far we’ve come:
Nov 2025 → Nov 2026
❌ A browser extension
✅ An open context layer that you control
❌ Context on a handful of major AI websites
✅ MCP support, so your context follows you into any tool or agent
❌ One big pile of memories
✅ Memory buckets for every project, client or team
❌ Hard to see what your AI knows
✅ Memory Studio, to see, edit and organize everything
❌ Changes you couldn’t undo
✅ Version history, so you can roll back anytime
❌ Just you
✅ Sharing with teammates and agents
Every one of these started with a user telling us what broke, what was missing, or what they wished their AI remembered.
So where do we go from here?
We believe work is changing. Three years from now, a typical team will be a mix of people, agents and tools.
Agents will do much of the work, and humans will set direction and step in when something needs to stop.
In that world, how humans and agents work together becomes the core problem. Teams will need a workspace where people can set the plan, steer, and close the gap between what they understand and what their agents actually do.
Today, that’s Memory Studio: a context hub that delivers the right context to the right agent. Tomorrow, it becomes the cockpit, where you see which agents are doing what, with which context, which permissions and which goals.
The plane will mostly fly itself. We’ll still be the pilots. That cockpit is what we’re building.
We’re launching on Product Hunt this Saturday. If this resonates, we’d love for you to come say hi. 👋


Replies
Be the first to reply
Have a question or a thought to share? Add a comment above to start the conversation.