Rillence

Agentic containers with dynamic UI.

Agent + plugins + UI. One container.

Install Ride

Every task in sight

Each task has its own tab with stages: done, running now, next. A mark tells whether it is working, waiting for you, stuck or paused, and why.

Subagents are listed on the tab of their task, each with what it is doing now. Drag a tab to change the priority.

Decide in a card

When an agent needs you, its question waits in a card with ready answers and a field for your own words. A ready answer only fills the field: nothing goes out until you press Send. Each task also takes an instruction, so you can steer it at any time.

All on one page

Full mirroring

One page on every device, every session in sync. A question or an answer on the phone shows on the laptop at once, and back again. Start on the phone, go on at the laptop, finish on the phone: nothing is lost. Many sessions at once come with Pro.

Chat

One chat with your agent, written in markdown: lists, tables, code. Nothing you wrote is ever lost.

Files

A file path in any text opens the file in a window: markdown rendered, video and audio in a player.

Limits

Claude and Codex limits sit at the bottom of the page, for five hours and for the week. A daily budget mark shows when today's share is spent.

Monitor

CPU, memory, temperature, every core and the busiest processes, on live graphs that barely load the machine.

Light, steady, open

Ride stays out of your way, so your attention goes to the agents, not to the tool. It is built on an open protocol: nothing hidden, nothing locked in.

  • 01LightOpens at once, even on a phone
  • 02One viewEvery screen shows the same tab
  • 03SteadyA restart loses nothing
  • 04OpenYours to build on

How you use it

  1. 01

    Agents work

    Your agents run on your machine and keep Ride up to date.

  2. 02

    You look in

    Open Ride on the laptop or the phone: the layout fits the screen, from a wide monitor to a folded phone.

  3. 03

    You decide

    Answer from wherever you are. The agent reads your answer and goes on.

Install

Ride ships as one Docker image for Linux x86_64. Your agent (Claude Code, Codex or another) starts it in a folder and reads what Ride puts there.

Image coming soon.

Paste this into your agent

Read RIDE.md and follow it

Native binary and MCP later.

Who it's for

Developers who keep AI agents at work all day.

Install Ride