Every AI coding session starts blind. Reset context or switch tools, and critical decisions and constraints vanish. MemHub provides a unified, local-first memory layer for Cursor, Claude Code, Windsurf, Cline & Codex: • Fast: <2ms reads/writes via local SQLite WAL. • Cloud Sync: Background sync in <700ms. • Secure: Zero code exfiltration; source code NEVER leaves your machine. • Portable: 1-click import & export to MEMORIES.md. Keep your context across tools without risking privacy.

Hey Product Hunt! 👋 We built MemHub because we were tired of our AI coding assistants repeating the same mistakes and losing track of architectural decisions every time context was compacted or reset. When teammates rely on different AI tools—Cursor for frontend, Claude Code for backend, Codex CLI in the terminal—project knowledge quickly becomes fragmented. MemHub fixes this by giving your agents a persistent, shared memory layer that runs entirely local-first and syncs across your team in real time. Get started in 30 seconds: npm install --global @simplexlat/memhub mh login mh init We’d love to hear your feedback, thoughts, and how your team currently manages context across AI tools! 🚀

MemHub

Persistent shared memory for AI coding agents

Persistent shared memory for AI coding agents