Skip to content
OMem

Install OMem

Read this if you just heard about OMem and want to try it. By the end of this page you’ll have the omem command installed and verified — about two minutes. The next three pages get you from “installed” to “my agent can answer questions about my work.”

OMem is available on macOS today; Windows is coming. It’s a single command-line tool — one binary, no Python to manage, no account to create.

The friendly path: let your agent install it

Section titled “The friendly path: let your agent install it”

If you already use an AI agent like Claude Code, the simplest install is to let it read the instructions and run them for you. Open your agent and say:

“Install OMem from github.com/seacen/omem and run me through the setup.”

It’ll do everything on this page and the next. If you’d rather do it by hand — or you want to understand each step — read on.

Install by hand

Section titled “Install by hand”

  1. Install the binary. One line:

    Terminal window
    curl -fsSL https://github.com/seacen/omem/releases/latest/download/install.sh | sh

    It installs omem to ~/.local/bin/omem — make sure that’s on your PATH.

  2. Verify it runs:

    Terminal window
    omem version

    You should see a version line print. That’s it — the tool is installed.

  3. Continue to setup to connect your data and pick an LLM.

$
Why it’s unsigned, and what that means for you

The macOS binary is unsigned (not notarized by Apple). The install script handles this for you: it downloads without the quarantine flag, so you won’t hit a Gatekeeper “unidentified developer” wall. This is a deliberate v1.0 choice (code-signing has real overhead for a solo project); nothing about it changes how OMem runs. If you ever download the binary some other way and macOS blocks it, removing the quarantine attribute (xattr -dr com.apple.quarantine /path/to/omem) clears it.

What’s next

Section titled “What’s next”

You have the tool; now connect it to your work.