Gemdex runs on your own machine — the client starts it, so there is no endpoint to ping. 66 installs a week from npm. Last commit 22 Oct 2026.
Persistent memory layer for AI coding agents: save, recall, update via Gemini plus LanceDB
What the code does
We read the source, 3 h ago · rules 995398eba69f
What this server is able to do. For an MCP server this is often the job itself — a terminal server runs commands because that is what it is for. Listed so you know what you are plugging in, not as an accusation.
resetProcess()
const child = spawn(command, args, {
const blobFiles = await fs.readdir(path.join(blobDir, memoryId), { recursive: true });
const blobFiles = await fs.readdir(path.join(blobDir, memoryId), { recursive: true });
Is this your server and something here is wrong? Tell us — corrections are free and do not require a plan.
This code can reach further than it looks
We found places where it runs commands, builds paths or queries from values it is given. None of that is a flaw by itself — it becomes one when the code changes, and code changes quietly between releases. We re-read it on every one.
Connect this server
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add gemdex -e 'GEMINI_API_KEY=<GEMINI_API_KEY>' -- npx -y gemdex-mcp{
"mcpServers": {
"gemdex": {
"command": "npx",
"args": [
"-y",
"gemdex-mcp"
],
"env": {
"GEMINI_API_KEY": "<GEMINI_API_KEY>"
}
}
}
}[mcp_servers.gemdex]
command = "npx"
args = ["-y", "gemdex-mcp"]
env = { GEMINI_API_KEY = "<GEMINI_API_KEY>" }{
"mcpServers": {
"gemdex": {
"command": "npx",
"args": [
"-y",
"gemdex-mcp"
],
"env": {
"GEMINI_API_KEY": "<GEMINI_API_KEY>"
}
}
}
}{
"servers": {
"gemdex": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"gemdex-mcp"
],
"env": {
"GEMINI_API_KEY": "<GEMINI_API_KEY>"
}
}
}
}Put your own value in place of
<GEMINI_API_KEY> (Google AI Studio key for local mode, or the exact value 'local' for on-device MLX text embeddings. Not needed in remote mode).
The author’s registry entry marks it as required.
Optional settings from the registry entry, left out of the block above:
GEMDEX_MODE (Memory backend: embedded local LanceDB or a self-hosted Gemdex Server), LANCEDB_PATH (Directory for the embedded LanceDB store), EMBEDDING_MODEL (Gemini text embedding model), GEMDEX_REMOTE_URL (Gemdex Server root URL (required in remote mode)), GEMDEX_REMOTE_TOKEN (Gemdex Server bearer token (required in remote mode)), GEMDEX_REMOTE_NAME (Human-readable name for the remote server), GEMDEX_TRUST_RANKING (Re-rank recall results by recorded worked/failed/stale outcomes).
Alternatives to Gemdex
same job, measured the same wayLocal-first persistent memory MCP server for AI coding agents.
Local-first persistent memory for AI coding agents.
Persistent memory layer for AI coding agents — your agent remembers what you taught it.
Local-first, persistent memory for AI coding agents — remember, recall, forget, and list, over MCP.
Persistent, verified memory for AI coding agents: recall, capture, guardrails, review queue.
Persistent, local-first memory for AI coding agents. Remember, recall and forget across sessions.
Shared, persistent memory for AI agents, with semantic recall and live updates.
Persistent memory for AI coding agents. Install the memry CLI and run memry setup.
Gemdex — questions
Answers built from our own checks of this server.