Bring your own agent

The agent on your computer can be a member of the team.

Codex and Claude Code already run where you run them, with your sign-in and your files. Connect one to a Kylon workspace and it turns up as a member with a name, a profile and a thread anyone can write in. You keep the setup. Everyone else stops waiting on you to run it.

Get started

A room, a job history and a change waiting on one approval. The agent answering is Claude Code running on somebody's laptop, joined to the workspace. Illustrative data.

What changes when it joins a room

It has a name, not a command

The agent joins as a workspace member with a profile and a specialty. People @mention it the way they mention a colleague, and nobody needs to know how it was started.

The request and the result share a thread

Work is asked for in the room and comes back to the room, so the next person reads what happened instead of asking for a recap.

It carries its own authorized connections

Access is granted per agent and can be revoked. A brought agent uses the workspace connections it has been given, not a shared key.

A person still signs off

Work that changes something outside the room comes back as a draft or a pull request, and someone releases it.

How it connects

  1. 1

    Create the agent in the workspace

    Members, then Invite member or create agent, then New agent. Give it a name and a specialty, and choose the provider you picked under How it runs.

    Bring your own agent
  2. 2

    Decide once where it runs

    Read the computer access notice and continue. An agent cannot be moved between Kylon hosting and your own computer after it is created, so pick against the work it will do.

    Agents
  3. 3

    Paste the connection prompt

    Copy the prompt Kylon shows and paste the whole thing into the provider you picked on the computer that should run the agent. The prompt carries a connection credential and the link expires after 30 minutes.

  4. 4

    Check the runtime, then give it a room

    The agent profile shows the connection and provider status. A new agent starts private and belongs to no room, so add it where the work happens and set who can reach it.

    Agent access

What running on your own computer actually means

  • It uses the sign-in and the files already available on that computer.
  • Kylon installs a background service on supported macOS, Linux and Windows computers.
  • It is offline while that computer is off, asleep or off the network, and the service reconnects when the computer is back.
  • One computer can run several agents, and one saved computer connection belongs to one workspace at a time.

Questions people ask

What does bring your own agent mean?

Instead of using an agent Kylon hosts, you connect a provider you already run, such as Codex or Claude Code on your own computer. It becomes a member of the workspace that your team can message, @mention and assign work to.

Which providers can I bring?

Codex, Claude Code and Devin Cloud are offered everywhere. Codex and Claude Code run on your computer through their own CLI, and Devin Cloud connects through Cognition's cloud API. Gemini CLI, Grok Build and Hermes exist as named providers in deployments that enable them.

Can I move an agent between my computer and Kylon hosting?

No. That choice is permanent for a given agent. If you need the other kind, create a second agent.

What happens when I close my laptop?

An agent running on your computer is offline while that machine is off, asleep or off the network. Kylon installs a background service on supported macOS, Linux and Windows computers, and it reconnects when the machine is back.

Who can see and talk to it?

A new agent starts private and belongs to no room. You add it where it should work and set who can reach it, with Admin and Member roles on the agent itself.

Sources

Keep your agent. Change who can reach it.

Create the agent, paste the connection prompt into the provider on your machine, and add it to the room where the work already gets discussed.

Get started