Changelog.

Every Tensor and Jarvis action now appears in a unified Receipts log with a timestamp, inputs, and a one-tap undo where applicable.

Orb memory consolidation runs every 90 seconds of idle time instead of on a fixed schedule. Multi-session recall improves measurably on long days.

Jarvis no longer trips its own wake word when reading output aloud.

Every Tensor answer now pins citations back to the exact DOM range on the source tab. Click to jump; the span highlights on arrival.

Fixed a race where Tensor synthesised answers before slow-loading tabs finished reading.

MingLLM now runs on saturday-4b, our unified 4B base tuned via 11 rounds of expert iteration. End-to-end browser task score: 94.9% on our internal 59-task set.

Inference latency reduced ~22% on M2/M3 Max. Time-to-first-token sits under 1s for most voice interactions.

Cloud calls are now off by default. MingLLM runs entirely on-device unless you explicitly grant a remote tool permission.

Added a privacy panel that enumerates every outgoing request from the last 30 days.

MingLLM preview opens on invite. Jarvis + Tensor + Orb memory.