Installation

Quick Install (curl)

Install with a single command:

curl -sSf https://hebbs.ai/install | sh

This installs the hebbs CLI, a lightweight client for connecting to an enterprise HEBBS server. It supports login, recall, remember, push, and workspace management over REST.

For the full engine (local vault mode with indexing, daemon, and Memory Palace):

curl -sSf https://hebbs.ai/install | sh -s -- --full

Add the install directory to your PATH if prompted:

export PATH="$HOME/.hebbs/bin:$PATH"

Homebrew (macOS & Linux)

brew install hebbs-ai/tap/hebbs        # CLI (connects to enterprise server)
brew install hebbs-ai/tap/hebbs-full   # Full engine (local vault mode)

Verify:

hebbs --version

Which Install Do I Need?

Install	What you get	Use case
Default (hebbs)	REST client CLI: login, recall, remember, push, workspaces	Enterprise developers connecting to a shared server
Full (hebbs --full)	Complete engine: init, index, watch, serve, panel, reflect	Local mode, standalone use, self-hosted server

Enterprise Server

Deploy the HEBBS server on your infrastructure in one line:

curl -sSf https://hebbs.ai/server | OPENAI_API_KEY=sk-... sh

Then connect your CLI:

hebbs login --endpoint https://hebbs.company.com:8080 --api-key hb_live_sk_...

Install Options

Pin a specific version:

HEBBS_VERSION=v0.1.0 curl -sSf https://hebbs.ai/install | sh

Change the install directory:

HEBBS_INSTALL_DIR=/usr/local/bin curl -sSf https://hebbs.ai/install | sh

Supported Platforms

Platform	Architecture	Binary
macOS	Apple Silicon (arm64)	hebbs-macos-arm64
Linux	x86_64	hebbs-linux-x86_64
Linux	aarch64	hebbs-linux-aarch64

Docker

Run HEBBS in a container:

docker run -p 6380:6380 -p 6381:6381 ghcr.io/hebbs-ai/hebbs:latest

With persistent storage:

docker run -p 6380:6380 -p 6381:6381 \
  -v hebbs-data:/data \
  ghcr.io/hebbs-ai/hebbs:latest

Systemd (Bare Metal / VM)

On Linux servers, install HEBBS as a systemd service for automatic startup, restart-on-failure, and log management via journald:

curl -sSf https://hebbs.ai/install | sudo sh -s -- --with-systemd

This creates:

Path	Purpose
/usr/local/bin/hebbs-server	Server binary
/etc/hebbs/hebbs.toml	Configuration file
/etc/hebbs/hebbs.env	Environment overrides and API keys (mode 600)
/var/lib/hebbs	Data directory (owned by hebbs user)
/etc/systemd/system/hebbs-server.service	Systemd unit

Enable and start the service:

sudo systemctl enable --now hebbs-server

Common commands:

sudo systemctl status hebbs-server    # check status
sudo systemctl stop hebbs-server      # graceful shutdown (SIGTERM)
sudo systemctl restart hebbs-server   # restart after config change
journalctl -u hebbs-server -f         # stream logs

Edit /etc/hebbs/hebbs.toml for server configuration and /etc/hebbs/hebbs.env for API keys and environment overrides.

Build from Source

Prerequisites: Rust 1.75+, protoc (Protocol Buffers compiler).

git clone https://github.com/hebbs-ai/hebbs.git
cd hebbs
cargo build --release

Binaries are output to target/release/:

./target/release/hebbs-server --version
./target/release/hebbs-cli --version

What’s Next

• Quickstart — start the server and store your first memory in under 5 minutes
• Configuration — customize ports, storage, embedding models, reflection, and more