Drop it on a spare box, a NAS, or an old laptop. You get a local large language model, a terminal assistant, and a ready-made set of automation scripts — nothing leaves your network, and you keep full control.
A small, self-contained install. No subscriptions, no lock-in, no telemetry.
An open-weight model served by Ollama on your own GPU or CPU — private, offline-capable, yours.
An interactive CLI that reads your project, edits files, and runs commands with your approval.
A starter set of ready-made scripts (status, memory, scheduling, backdrops) you can keep or swap.
Turn on a spoken voice, a voice-to-text pipeline, or a remote model box — only if you want them.
A documented, generic "memory" file so the assistant knows your conventions — never your private data.
Inference stays on your machine. Nothing is phoned home; there's no account to create.
Works out of the box; the add-ons are opt-in.
Download & run the installer. One command checks your machine, installs Ollama + the model + the CLI, and drops the scripts in place.
Answer a couple of questions. The onboarding asks whether you want a voice, a voice-to-text pipeline, or a separate model box — and sets up autostart.
Talk to it. Start the assistant, point it at a project, and it's working. Everything runs locally from here on.
One-time purchase · yours to keep, update, and tinker with
Lifetime downloads included.
Secure checkout via Lemon Squeezy. If the button isn't active yet, email [email protected] and we'll send the link within a day.
Short notes on local-AI setups, new packs, and lessons learned. One email a week, tops.