Gemdex MCP server

A global, persistent memory layer for AI coding agents — Gemini embeddings × LanceDB × MCP

⭐ Star on GitHub · 📦 npm · 💬 Discussions · 🐛 Issues

Why Gemdex

Your agent re-learns everything every session. You explained your deploy flow

last week; today it has no idea. Gemdex gives it durable memory you write

on purpose — once, recallable everywhere.

A memory layer is a deliberately written, persistent store that is the source of truth. You teach your agent something once, and it remembers forever, across every repo and every session.

update_memory. No silent capture, no background recall.

• 🧠 You decide what to remember — explicit save_memory / recall /

scopes, no folders, no tags; embeddings do the disambiguation.

• 🌍 One global pool — every memory is searchable from everywhere. No

chunks, but recall always returns the full memory, never a fragment.

• 🔎 Sharp recall, whole answers — hybrid semantic + BM25 over internal

(Claude Code, Cursor, Codex CLI, Windsurf, Cline, Continue, Zed…) works instantly.

• 🔌 Plug-and-play — speaks MCP over stdio, so any compatible client

LanceDB at ~/.gemdex, or run the whole stack on infrastructure you own with one command and share one memory pool across every machine.

• 🪶 Local by default, self-hosted when you want more — start with embedded

Google login) to browse / edit / delete / export / import, and to upload chat transcripts.

• 🌐 Browser manager — the self-hosted stack ships a web UI (behind your

The motivating workflows

During a session:
  "We just figured out how to wire up the Junie review workflow — save that to memory."

Weeks later, a different repo:
  "Set up the Junie review workflow here — check your memory layer for how we do it."

Different machine, different app:
  "Notarize and sign this build — the credentials and steps are in my memory layer."

Two ways to run it

Shortened. The full README is on GitHub.