A nostalgic desktop assistant with a high‑fidelity Clippy and on‑device AI chat powered by node‑llama‑cpp. The app lives in the clippy-ai/ folder.
clippy-ai/— Electron app source, scripts, and docsProductRequirements/— PRD and requirements docs
# From repo root
cd clippy-ai
# 1) Install dependencies
npm install
# 2) Download a starter model (if models/ is empty)
npm run download-model
# 3) Start the app in dev mode
npm startcd clippy-ai
npm run build:win # Windows
# or on macOS/Linux
npm run build:mac
npm run build:linuxThe build auto‑fetches a default model if none exists and bundles models/ as an extra resource.
- Models: place
.gguffiles inclippy-ai/models/and pick them in Settings. - macOS builds may require signing/notarization for distribution.
MIT
