Skip to the collection

agent-native · Agent Runtime & Infrastructure · Verified 2026-09-26

OpenAI Symphony

Listed on The Agent-Native Index

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

What it does

Symphony continuously reads eligible work from an issue tracker, claims an issue, creates a deterministic isolated workspace, and launches a Codex app-server session inside it. It keeps the agent working through bounded turns, reconciles tracker state, schedules retries, and exposes structured runtime status. The work policy lives in version-controlled WORKFLOW.md, so the prompt, completion rules, proof requirements, and handoff state travel with the codebase.

The language-neutral SPEC.md is the product contract. The included Elixir/OTP implementation is a reference implementation and supports Linear, GitHub Issues, Jira Cloud, Asana, and GitLab.


Agent-facing interfaces

How agents onboard

Interaction pattern: Agent-readable specification or Elixir reference service

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

See the full onboarding section in the dossier.

Protocol surface

Interface Detail
Specification SPEC.md — language-neutral normative service contract
Agent runtime Codex app-server protocol
Tracker integrations Linear, GitHub Issues, Jira Cloud, Asana, GitLab in the reference
Configuration YAML front matter plus prompt body in repository-owned WORKFLOW.md
Operations Structured logs, optional JSON API/dashboard, workspace lifecycle hooks

Verification

Last verified 2026-09-26.

Sources: