Symphony is an open-source coding agent orchestrator and harness developed by OpenAI. Built on Elixir and the Erlang VM (BEAM), Symphony provides the infrastructure required to manage, execute, and monitor autonomous coding agents in isolated development environments. In practical workflows, Symphony orchestrates coding agents across dedicated git worktrees and tasks. It handles task dispatching, skill execution, and lifecycle management while interfacing with issue trackers and code hosts such as GitHub and GitLab. The system manages tool execution and runbooks through structured repo-local skills and automated checks. Symphony distinguishes itself through its host-boundary security model and process isolation. It enforces strict credential separation, actively scrubbing host git and tracker authentication token aliases (such as GitHub and GitLab credentials) from agent execution runtimes to prevent autonomous agents from accidentally leaking or abusing host-level privileges. Symphony is open source, distributed via GitHub, and packaged as self-contained executables via Burrito alongside standard Elixir/Mix build pipelines.
Host-Agent Credential Isolation — Enforces strict execution boundaries by scrubbing forge authentication tokens (GitHub/GitLab) from agent execution environments. - BEAM-Powered Orchestration — Built in Elixir to leverage OTP concurrency and fault tolerance for reliable multi-agent task execution. - Repo-Local Skills Integration — Coordinates reusable agent procedures, validation routines, and release runbooks via structured skill definitions. - Git Worktree Isolation — Runs agent tasks in segregated worktrees to prevent workspace collisions and uncommitted state pollution. - Standalone Burrito Binaries — Distributes portable, single-binary releases across target architectures without requiring local Erlang/Elixir toolchains. - Automated Verification Pipelines — Integrates repo-level test suites, linting checks, and smoke tests into agent landing loops.
Use cases
Use Case: Automated Issue Resolution A developer assigns incoming repository issues to Symphony, which spins up an isolated worktree, launches a coding agent with sandboxed credentials, applies the necessary code changes, runs the project test suite, and opens a pull request.
Use Case: Safe CI/CD Agent Execution A team runs Symphony within their automation pipelines to execute repo maintenance tasks while ensuring that agent processes cannot access or exfiltrate pipeline secrets or elevated GitHub tokens.
Use Case: Release and Task Skill Execution Maintainers run standardized repository procedures—such as version bumping, change verification, and tag generation—by invoking predefined repo-local skills orchestrated through Symphony.
Who it is for
Symphony is built for platform engineers, AI tooling developers, and software teams looking for a secure, isolated orchestration harness to manage autonomous coding agents. It is less suited for developers seeking a simple interactive chat agent or direct IDE autocomplete extension.
Install guide
Installation To build Symphony from source, ensure you have Elixir, Erlang/OTP, and standard build tools installed: bash git clone https://github.com/openai/symphony.git cd symphony/elixir mix deps.get make all Alternatively, download the precompiled binary from the repository's GitHub Releases page once artifacts are published.
First Run Verify that the environment dependencies and forge CLI tools (such as gh) are configured without exposing ambient forge credentials directly to agent child processes: bash # Verify build and environment readiness make -C elixir all
Example Session Run Symphony skills or launch orchestrated agent runs against your project repository: bash # Execute verification checks across the codebase mix test # Run Symphony task automation and skill validators python3 .codex/skills/.system/skill-creator/scripts/quick_validate.py .codex/skills/release
SWE-agent — SWE-agent is an open-source autonomous software engineering agent harness built and maintained by researchers from Princeton University and Stanford University.…
Agent Runbook — Agent Runbook is an open-source contract-based multi-agent skill framework developed by KnoxOps. It acts as an orchestrator and compiler that…
Weaver — Weaver is an open-source agent harness and orchestration framework maintained by sean35mm on GitHub. Built on top of the Bun…
Context Bridge — Context Bridge is an open-source CLI agent harness and context orchestration utility developed by serdardb. It is built to facilitate…
Hivelore — Hivelore is an open-source agent harness and contextual memory orchestrator created by Doucs91. It provides persistent project memory, decision tracking,…
Agentic Engineering Framework — Agentic Engineering Framework is an open-source orchestration harness and governance layer developed by Dimitri Geelen. It is designed to constrain,…