Levh — Local-first memory for AI agents and workflows (launching soon on PH!)
by•
Hey everyone! We're building Levh, a local-first memory layer for AI agents and workflows. It captures context from conversations, projects, files, email, calendars, and MCP tools, builds an inspectable knowledge graph, surfaces trust and conflict signals, and lets stale details fade over time — all without cloud accounts or telemetry.
We're launching on Product Hunt soon and would love your early feedback, thoughts, or questions about the idea before we go live. Thanks for checking it out!
6 views
Replies
As a vibecoder, I use several different AI agents, but since there's no shared memory between them, I kept running into context loss and wasted time. Files like AGENTS.md or MEMORY.md just weren't enough on their own.
I've been working on this for about 6 months now (starting over from scratch twice along the way), and Levh is finally ready to go. I really hope it turns out useful for others dealing with the same problem.
Feel free to check it out and try it — any feedback, bug reports, or suggestions would mean a lot. You can use it as an MCP, and connect it with whatever other tools you're already using.
We're actually launching officially tomorrow, so I'll share that here too once it's live. Thanks for taking a look!