Getting Started

Digital Me carries your intelligence — knowledge, taste, decisions — to every agent you run. This page takes you from a clean machine to your first capture → recall loop in about five minutes.

Prerequisites

• Node.js ≥ 22.5 — to install and run the CLI from npm. (pnpm is only needed for the From source path below.)
• A Gemini API key — memory_search embeds your wiki with gemini-embedding-001 and the nightly dream-cycle uses the same key. You add it to one file right after setup.
• Python ≥ 3.11 — optional; only needed for the dream-cycle distillation pipeline (skip it with --minimal).

That's the whole list. The brain — brain-host — ships with the CLI and setup installs it as an always-on service; there is no gateway daemon to install first. openclaw is one of the supported runtimes, not a prerequisite.

Install

# Install the CLI from npm
npm install -g digital-me

# Detect installed CLIs, scaffold ~/digital-me/, install hooks/skills/configs,
# wire every runtime, run doctor:
digital-me setup

# Node-only? Skip the optional services (dream-cycle venv + dashboard build):
digital-me setup --minimal

# Pin a custom wiki location:
digital-me setup --wiki-root ~/notes/brain

Then give the brain its key — one line in ~/digital-me/.data/.env, which brain-host loads and every nightly worker inherits:

echo 'GEMINI_API_KEY=...' >> ~/digital-me/.data/.env
digital-me doctor        # the key is seen, the brain is serving

What setup does:

1. Detects ~/.claude/, ~/.codex/, ~/.hermes/, ~/.openclaw/ to figure out which runtimes you have.
2. Scaffolds the wiki root — ~/digital-me/{wiki,tastes,inbox,.cache,.data} plus a live config.yaml (created only if absent, never clobbered) and a .gitignore that keeps .data/ out of your wiki repo.
3. Installs brain-host — the hub: links the service, writes its bearer token, builds the retrieval index over wiki/ + tastes/, and loads it as a launchd / systemd --user service on 127.0.0.1:18791.
4. Installs each detected runtime, pointed at brain-host — hooks, skills, MCP entries, protocol bundles (see Runtimes for what lands where).
5. Runs doctor to confirm everything resolved.

Re-running is idempotent — installers merge into existing settings without clobbering your other hooks.

From source

For contributors, or to run an unreleased version: clone and build, then drive the CLI with pnpm dm <command> from the repo — that's the in-repo equivalent of the digital-me command before it's linked onto your PATH.

git clone https://github.com/Amyssjj/digital-me.git ~/digital-me-os
cd ~/digital-me-os && pnpm install && pnpm build
pnpm dm setup        # `pnpm dm` = the CLI, run from the clone

Packaging: the CLI installs from npm as digital-me; during setup it pip-installs the dream-cycle pipeline from PyPI (digital-me-dream-cycle).

Your first loop

The whole system exists for one loop: an agent learns something once, and every agent knows it afterwards.

# 1. Verify the brain is serving
digital-me doctor
digital-me service brain-host status

# 2. In any wired agent (Claude Code, Codex, Hermes, openclaw), do real work.
#    When something reusable surfaces, the agent calls learning_capture —
#    or you ask it directly: "remember this for next time".

# 3. Watch it come back: start a NEW session, ask about the same topic.
#    The runtime adapter injects matching wiki knowledge into the prompt
#    before the model ever sees your question.

Behind the scenes the capture lands in your wiki repo as a reviewable markdown entry, the nightly dream-cycle distills raw learnings into clean entries, and memory_search serves them to every runtime. Nothing is hidden in a vector store you can't read — the wiki is plain files in git.

Where things live

With the npm install, the only directory that's yours to own is ~/digital-me — the CLI itself lives in npm's global packages.

(Installed from source? The repo checkout lives at ~/digital-me-os — code only, no personal data.)

Next steps

• Wire up each agent CLI you use → Runtimes
• Understand the loop you just ran → How it works
• Every command and flag → CLI reference