# openai/symphony

Symphony turns project work into isolated, autonomous implementation runs, allowing teams to manage work instead of supervising coding agents.

Repository: https://github.com/openai/symphony
Canonical: https://ross.abutalabs.com/products/symphony
Homepage: https://openai.com/index/open-source-codex-orchestration-symphony/
Language: Elixir
License: Apache-2.0
License Family: permissive
Last push: 2026-08-19T23:24:27+00:00

## Health v2 (maintenance only)
Score: 80/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 94, longevity 13
- inputs: {"age_days": 188, "days_push": 14, "days_rel": 40, "gap_med": 6, "n_releases_24m": 2}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 26871, forks 2755 (observed 2026-08-28T04:11:47.164079+00:00)

## What it is
Symphony is an open-source spec and Elixir reference implementation from OpenAI that turns a project-management board like Linear into a control plane for autonomous coding agents such as Codex. Each open task gets an agent that runs continuously and produces proof of work (CI status, PR review feedback, complexity analysis), while humans review and accept the resulting pull requests at a higher level.

## Use cases
- orchestrate multiple coding agents across a backlog
- turn Linear issues into pull requests automatically
- run Codex agents autonomously on project tasks
- reduce context switching between coding agent sessions
- manage a fleet of AI coding agents from a project board
- auto-assign agents to tasks and land PRs safely
- scale agentic software development across a team

## When to choose
- Your codebase already follows harness engineering practices with strong automated tests and guardrails
- Your team uses Linear or a similar board and wants tasks dispatched to coding agents automatically
- You want engineers to review finished work (PRs with CI status and proof of work) instead of supervising individual agent sessions

## When to avoid
- You operate in an untrusted or high-risk environment - it is an engineering preview intended for trusted environments
- Your repository lacks CI, automated tests, and guardrails to validate agent-generated code
- You need tight human supervision of every agent step rather than autonomous implementation runs

## Facets
- artifact type: application
- maturity: experimental
- function: agent-framework, workflow-automation, developer-tools
- domain: developer-tools
- platform: self-hosted
- tags: codex, openai, coding-agents, agent-orchestration, linear, issue-tracker, autonomous-agents, pull-request-automation, harness-engineering, spec, elixir, ai-agents, automation, devops, linux, macos

## Member repositories
- openai/symphony (main) score 80

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:47.164079+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:55:49.016942+00:00, confidence not recorded.
  - readme: https://github.com/openai/symphony (fetched 2026-08-28T04:11:47.164079+00:00, sha b47702596a9f)
  - homepage: https://openai.com/index/open-source-codex-orchestration-symphony/ (fetched 2026-08-29T07:52:00.568826+00:00, sha 51d15d42b3ec)
  - site_page: https://openai.com/about (fetched 2026-08-29T07:52:00.590013+00:00, sha de4b627b33e3)
  - site_page: https://developers.openai.com/api/docs (fetched 2026-08-29T07:52:00.592377+00:00, sha d617c5215f2a)
  - site_page: https://developers.openai.com/ (fetched 2026-08-29T07:52:00.594545+00:00, sha 7d240c906018)
- Data as of 2026-08-30T08:39:29.467469+00:00.