openai/symphony

What it is

Symphony turns project work into isolated, autonomous implementation runs, so teams can manage work instead of supervising coding agents. In the demo, it monitors a Linear board for work and spawns agents to handle tasks. The agents provide proof of work such as CI status, PR review feedback, complexity analysis, and walkthrough videos, and land the PR safely when accepted. It is described as a low-key engineering preview for testing in trusted environments.

Who it's for

• Teams that have adopted harness engineering and want to manage work at a higher level than individual coding agents
• Developers who want to implement Symphony from a spec in a language of their choice
• Engineers interested in trying an experimental Elixir-based reference implementation

Requirements

Setup

1. Option 1: Build your own implementation

   Tell your favorite coding agent to build Symphony in a programming language of your choice using the spec.

   bashCopy

   Implement Symphony according to the following spec:
   https://github.com/openai/symphony/blob/main/SPEC.md

2. Option 2: Use the experimental Elixir reference implementation

   Follow elixir/README.md to set up your environment and run the Elixir-based implementation. You can also ask a coding agent to help with setup.

   bashCopy

   Set up Symphony for my repository based on
   https://github.com/openai/symphony/blob/main/elixir/README.md

Examples

Implement Symphony according to the following spec:
https://github.com/openai/symphony/blob/main/SPEC.md

Set up Symphony for my repository based on
https://github.com/openai/symphony/blob/main/elixir/README.md

Pros & cons

Images