Launching today
AO2 Memory
Portable context for every AI. Never explain yourself twice.
18 followers
Portable context for every AI. Never explain yourself twice.
18 followers
You explain your project to one AI. Then you open another and explain it all over again — none of them share a memory. AO2 Memory is a shared memory layer. It's an MCP server you connect to whatever you use — Claude, ChatGPT, OpenClaw — so they all read and write to the same place. Tell one agent something, and the rest already know. And you stay in control: see what every agent did, roll back any change, and revoke access anytime.
Open Notebook
@nazar_ilamanov Congrats on the launch! 🎉 A shared memory layer across different AI tools sounds really useful. Does AO2 Memory work with locally hosted models as well, or only cloud-based agents?
Open Notebook
@nicole_hynek Thank you Nicole! It works with anything! If your local interface supports MCP, you can connect via MCP or if not, just use the API directly.
Congrats on the launch! 🚀
A shared memory layer across AI tools solves a problem I've run into many times.
I'm curious: if two agents update the same piece of memory with conflicting information, how does AO2 Memory resolve it? Does it support versioning, conflict detection, or require manual approval before merging changes?
Open Notebook
@prashant_patil14 thank you Prashant! In this case the latest change will be the most up to date. But every record has a version history so the previous change will still be visible and you would be able to revert to it!
Mailwarm
How do you prevent one agent from overwriting memory, like namespaces?
Open Notebook
@karimbenkeroum so the latest updates take precedence but each record keeps a version history so you can roll back to any past version!