Skip to content
This repository was archived by the owner on Nov 9, 2026. It is now read-only.

v0.3.0

Choose a tag to compare

@speedyrulz speedyrulz released this 28 Oct 21:31
· 7 commits to master since this release

Model settings from your phone, and more reliable notifications.

Needs RulZ Coding Agent 0.8.0 or later on the computer for Model settings (the rest works with 0.6.0 and later). The GPU split, draft tuning and memory estimate also need RulZ LLM Suite 0.3.7 or later.

What's new

  • Model quick switch: tap the model under the project name to switch models, with a shortcut to Model settings.
  • Settings > AI model > Model settings, the same as on the computer:
    • How it writes: temperature, top P, top K, repeat penalty, longest answer, and thinking (on/off, reasoning effort, thinking budget and its message), saved per model.
    • How it loads: context length, GPU layers, flash attention, KV cache types, GPU split (all, one GPU, or a slider per GPU), draft (MTP) tuning, and RulZ's memory estimate as you change them. Saved in RulZ LLM Suite; a loaded model restarts with them.
  • Notifications: shown through the app's own background service so they arrive reliably; the ongoing notification says what the agent is doing (working, waiting for your OK, a question, paused, reconnecting); tapping one opens the project; Settings explains why notifications can't show and shows the last one sent.

Install

Download RulZ-Agent-0.3.0.apk on the phone, open it, and allow installing from that source. It installs over 0.2.0 and keeps your paired computers.

This is a debug-signed build (Android 7.0 and newer).