Store and retrieve persistent AI memory across tools using local SQLite with zero external dependencies. Built for developers who need portable, cross-tool context that survives between sessions.
io.github.sparkvibe-io/memorymesh
Local install
STDIO
No auth required
How models use it and what it is built for.
Store and retrieve persistent AI memory across tools using local SQLite with zero external dependencies. Built for developers who need portable, cross-tool context that survives between sessions.
Local install — runs as a subprocess.
Where to find authoritative docs and source for Sparkvibe IO Memorymesh.
Paste any of these into Agent Studio after connecting Sparkvibe IO Memorymesh.
Common questions about connecting and running Sparkvibe IO Memorymesh.
What is MemoryMesh and why would I use it?
MemoryMesh is a persistent memory layer for AI that runs locally on SQLite with no external dependencies. It lets you store conversation context, decisions, and data that persists across sessions and can be used by multiple tools.
Does MemoryMesh require internet or external services?
No. MemoryMesh is local-first and zero-dependency—it uses only SQLite, which runs entirely on your machine. All data stays local unless you explicitly export it.
Can I use MemoryMesh with multiple tools at once?
Yes. MemoryMesh is designed to be cross-tool portable, so you can connect the same memory database to different applications and AI tools via the MCP protocol.
How do I set up MemoryMesh?
Install it with `uvx memorymesh@4.1.0`. It communicates via stdio transport, so you connect it to your MCP client (like Claude Desktop or a custom integration) and it will create and manage a local SQLite database automatically.
What happens to my memories if I uninstall MemoryMesh?
Your SQLite database file remains on disk. You can reconnect to it by reinstalling MemoryMesh or by querying the database directly with any SQLite tool, since there are no proprietary formats or dependencies.
MCP Playground runs 10,000+ hosted MCP servers — GitHub, Linear, Notion, Stripe, Sentry and more — across Claude, GPT, Gemini, DeepSeek and 30+ AI models. Compare model answers side-by-side, save agent presets, share runs. Zero install.
Open Agent Studio