Open source

Technical information

For the person who installs Albiex Synth City and keeps it running.

What you need

Synth City must run in the same environment as your agent tools and the sign-ins they use. Do not run Synth City in Docker; Docker is not supported.

You need:

  • Python 3.11 or newer
  • Node.js 22.12 or newer
  • At least one supported agent tool installed in that same environment

The Python and Node.js versions above are general minimums. A tool you choose may need a newer version or extra preparation, so check its notes in the installation guide.

Install Synth City

At a glance:

  1. Get a copy of Synth City's source code.
  2. Run the install script. It prepares what Synth City needs to run.
  3. Start Synth City, then open it in your browser.
  4. Run the checks in the guide to confirm the installation.

The full guide gives the exact commands for each step.

Read the installation guide

Connect your AI harnesses and models

A harness is the agent tool a team member runs. Synth City works with these today, with more on the way. Each one runs signed in with your own subscriptions.

  • Claude Code
  • Codex
  • GitHub Copilot
  • OpenCode
  • Pi
  • Muse Code

Install a harness in the same environment as Synth City first. Then choose the harness and AI model for each role on your AI teams.

The harness guide explains what Synth City can show and set up for each choice.

Read the harness guide

See how your agents work

Two features show you how your AI teams operate. Agent Mail keeps every message the team members send each other: questions, context and handovers. Agent Bridge shows each agent's live session, with a live preview of its work.

Read the Agent Mail guideRead the Agent Bridge guide

Turn on your AI teams

Autonomy stays off until you turn it on for a team. To get ready:

  1. Give Synth City access to GitHub from the environment where it runs.
  2. Prepare each team roster, choose its AI Leader, and launch the Leader as the guide explains.
  3. Choose the GitHub repository Synth City should watch.
  4. Mark your features ready in GitHub when they can start.
  5. Choose how finished work may merge, and set finite limits for automated work.

When the rosters and watched repositories are ready, turn autonomy on. You can pause a watched repository separately.

Read the autonomy guideRead the Teams guide

Where to go next

Start at the docs home for the full guides. For the source code, choose “View product source” there.