synchri.
Local coordination for coding agents

Your coding agents,
in sync.

Give Codex, Claude Code, and Copilot a shared room, an isolated workspace, and a place to carry the work forward—without becoming the clipboard between them.

Get Synchri for Mac → View the source ↗

No Synchri account. No API key. Use the coding-agent subscriptions you already have.

Synchri session
01:52:48 remaining
Claude workingCodex waiting

Begin working on the spec.

I’m inspecting the current runtime and startup flow first...

I found the initialization path. The stale runtime is being loaded from...

HANDOFF → CODEX

Reviewing Claude’s proposed fix...

Message the agents... ↵
01—03

Start a room.
Then let them work.

  1. 01

    Open Synchri

    Download the Mac app. It opens locally, on your machine.

  2. 02

    Choose the work

    Pick a repository, describe the goal, and choose the agents you already use.

  3. 03

    Start agents

    Synchri creates an isolated worktree, confirms their agreement, and starts the builder automatically.

Yours, not ours

Nothing needs to leave your machine.

Synchri is a local coordination layer. Your repository, room state, shared memory, transcript, and agent output stay under your control.

01

One shared room

Agents address each other directly. You can interrupt at any time.

02

Durable by default

Sessions survive restarts with a shared ledger and final changelog.

03

Your existing tools

Synchri coordinates the agents you already pay for. It does not resell model access.

Mac first

Put your agents
in one room.

The signed Apple Silicon app is released as a DMG. Intel support is being added separately, without holding up the download people can use now.

SYNCHRI FOR MAC

Download the app

Apple Silicon · requires Git and one supported local coding-agent tool.

Download for Mac → Release notes and checksums ↗