How do you handle memory for your AI tools? We built an open-source solution.
Hey Product Hunt! 👋
We've been frustrated by the same problem most AI users face — every conversation starts from zero. Your AI doesn't remember your preferences, your projects, or what you told it yesterday.
So we built PersistMemory — an open-source memory layer that connects to ChatGPT, Cursor, VS Code, Claude, and Windsurf via MCP (Model Context Protocol).
How it works:
Connect PersistMemory to any AI tool that supports MCP
Your conversations automatically extract and store key facts, preferences, and context
Next time you chat, your AI already knows your tech stack, coding style, project details — everything
What makes it different:
🔓 Open source — self-host or use our cloud
🧠 Auto fact extraction with knowledge graph
🔗 Works across multiple AI tools (not locked to one)
🔐 End-to-end encryption support
⚡ Built on Cloudflare Workers — fast globally
Curious to hear from the community:
How are you currently handling AI memory/context across sessions?
What's the #1 thing you wish your AI remembered about you?
Would you prefer self-hosted or cloud for something like this?
Would love your feedback and thoughts. Drop a comment or try it out — link in our profile!
Replies