OpenEng
OpenEng is a local, on-device agentic CLI. Every prompt runs a
four-step pipeline — categorize, plan, execute, summarize — with
embedded models running in-process via llama.cpp. Threads, memory,
and models stay on device; there is no telemetry. The CLI is live
today on Linux x86_64 (install with: curl -fsSL
https://openeng.ai/cli/install.sh | bash); macOS and Windows are
coming soon. Visit openeng.ai in a
browser with JavaScript enabled to see the full site.