Provides a private memory layer for AI assistants that stores and recalls information locally on your computer.
Also known as ModelBrain/modelbrain-mcp, modelbrain-mcp
- Stars
- 0
- Tools
- —
- Downloads, 30 days
- —
- Latest version
- 0.1.5
- Last commit
- Nov 10, 2026
- License
- MIT
What is ModelBrain MCP?
ModelBrain provides a private memory layer for AI assistants that stores and recalls information locally on your computer.
Summary generated from the project's README
About ModelBrain MCP server
Giving AI assistants a private memory layer that stays on your local machine is what modelbrain-mcp enables. It acts as a thin stdio client that connects to the ModelBrain desktop application already running on your computer. This setup allows different assistants to share a single memory source while keeping all data stored locally rather than in the cloud.
Developers use this package to bridge MCP clients with the ModelBrain daemon. While Claude Desktop, Claude Code, and Cursor users can register the app directly, other clients can connect using npx. If you enable client tokens in the ModelBrain settings, you must provide a token via the MODELBRAIN_CLIENT_TOKEN environment variable. The server provides access to fourteen specific tools including remember, recall, list_memories, expand, forget, and various partition tools to manage your stored information.
What you can do with it
- 1Retrieve past contextInstruct an AI to look up previously saved information from your local memory layer.
- 2Manage memory contentsTell your assistant to delete specific pieces of information or organize them into different categories.
Written from the project's README. Every use case is checked against a quote from it.
How to install ModelBrain MCP
Formats verified 10 Nov 2026Install ModelBrain MCP in Claude Code
$ claude mcp add modelbrain-mcp -- npx -y modelbrain-mcpInstall ModelBrain MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"modelbrain-mcp": {
"command": "npx",
"args": [
"-y",
"modelbrain-mcp"
]
}
}
}Install ModelBrain MCP in VS Code
.vscode/mcp.json
{
"servers": {
"modelbrain-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"modelbrain-mcp"
]
}
}
}Install ModelBrain MCP in Claude Desktop
claude_desktop_config.json
{
"mcpServers": {
"modelbrain-mcp": {
"command": "npx",
"args": [
"-y",
"modelbrain-mcp"
]
}
}
}Install ModelBrain MCP in Codex
~/.codex/config.toml
[mcp_servers.modelbrain-mcp] command = "npx" args = ["-y", "modelbrain-mcp"]
Install ModelBrain MCP in Gemini CLI
~/.gemini/settings.json
{
"mcpServers": {
"modelbrain-mcp": {
"command": "npx",
"args": [
"-y",
"modelbrain-mcp"
]
}
}
}Automated checks
Ran Nov 14, 2026Full report →Passed our automated checks. That is not a security audit: read the code and give the server the narrowest access it needs. How checks work
- Source repository linked and reachablemodelbrain/modelbrain-mcp is public and reachablegithub.com/modelbrain/modelbrain-mcp ↗Pass
- Package names this repository as its sourcemodelbrain-mcp names modelbrain/modelbrain-mcp as sourcenpmjs.com/package/modelbrain-mcp ↗Pass
- Build provenance (Sigstore attestation)No build attestation published for this releaseNot checked
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-11-11github.com/modelbrain/modelbrain-mcp ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNo popular package one edit away from modelbrain-mcpPass
- No critical advisory in the package or its dependenciesthe package checked; no critical advisoryosv.dev ↗Pass
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeHTTPS endpoint answered 401 (sign-in required)Pass
README
View on GitHubmodelbrain-mcp
A thin MCP stdio client for ModelBrain: the private memory layer for AI assistants: one memory for every assistant you use, kept on your own computer.
This package does not install or bundle ModelBrain. It connects to the ModelBrain desktop app already running on your machine.
If ModelBrain is not installed
The launcher still starts and answers the MCP handshake, so your client shows the server and its tools instead of a start-up failure. Every tool call then returns one plain result: ModelBrain is not installed, and where to get it. Nothing is stored or recalled in this state. The same message is printed to standard error.
The tool list it shows comes from bin/tools.json, a copy of what the real server reports. Maintainers refresh it with npm run snapshot-tools on a machine that has ModelBrain installed.
To get the old behaviour, where the launcher exits with status 1 when the app is missing, pass --require-app or set MODELBRAIN_MCP_REQUIRE_APP=1.
Why a separate package
ModelBrain's own installed app already ships a binary, mcp_server, that speaks MCP over stdio and talks to the local ModelBrain daemon. This package finds that binary on the current machine and runs it. It exists so ModelBrain can be registered with the MCP Registry and the directories that list from it, with a reviewable package as the distribution target, while the actual engine stays out of this repository.
Requirements
- macOS (Apple Silicon), Windows (x64 or arm64) or Linux (x86_64 or aarch64, glibc 2.28 or newer)
- ModelBrain installed
Use
In ModelBrain's Allow or Block prompt this client shows up as your assistant's name with node as the program (mcp_server.exe on Windows). If you turn on Require client tokens in ModelBrain, mint a token for your assistant with modelbrain_admin setup tokens --mint NAME and put it in your client's env block as MODELBRAIN_CLIENT_TOKEN; the launcher passes it through.
Most MCP clients run a server with npx. Point your client's MCP config at:
{
"mcpServers": {
"modelbrain": {
"command": "npx",
"args": ["-y", "modelbrain-mcp"]
}
}
}
Claude Desktop, Claude Code and Cursor users do not need this: ModelBrain's own app registers itself with those clients directly. See modelbrain.net/docs for the full tool surface (14 tools: remember, recall, list_memories, expand, forget, six partition tools, diagnostics, usage, ping) and the connect guides for other clients.
License
MIT, for this wrapper package. ModelBrain itself is separate, proprietary software; see modelbrain.net for its own terms.
Security
See SECURITY.md.
ModelBrain MCP questions
How do I add ModelBrain MCP to Claude Code?
Run claude mcp add modelbrain-mcp -- npx -y modelbrain-mcp in your terminal. Add --scope user to make it available in every project.
How do I add ModelBrain MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
Do I need to install anything extra if I use Claude Desktop or Cursor?
No, ModelBrain's own app registers itself directly with Claude Desktop, Claude Code, and Cursor.
What happens if I run the server without having ModelBrain installed?
The launcher will start and show the tools, but every tool call will return a message stating that ModelBrain is not installed.
How can I secure the connection with client tokens?
You can mint a token using the modelbrain_admin tool and add it to your client's environment block as MODELBRAIN_CLIENT_TOKEN.