StackMap
Subscribe
Explore / harnessrouter
HarnessRouter

harnessrouter

Self-hosted Apache-2.0 reference implementation of the Unified Harness Protocol: one OpenAI-Responses-style API running Codex, Claude Code, Hermes, Pi and DSH with sessions, streaming, files, traces.

1,129 112 Python Apache-2.0updated yesterday
View on GitHubDispute this mapping →
Curator's take

Use it when your *product* needs to run coding-agent harnesses as a backend service and you don't want an integration per harness: one Docker image ships Console + Gateway + Runner, installs the harness CLIs you enable, and exposes runs/sessions/artifacts through UHP so you can swap Codex for Claude Code (or a cheaper harness×model pair — their benchmarks show 90%+ cost spread) without touching your app. Bring your own provider keys. It is a server for embedding harnesses, not a desktop cockpit for a developer (see agent-orchestrator, helmor) and not a model gateway (put litellm/omniroute underneath). Community Edition of a commercial product; check what stays in the paid tier before betting on it.

Mapped by ShipWithAI editors · links verified

Continue your stack

What teams reach for next — and why each earns a place beside harnessrouter. Ranked by curator confidence.

alternativealternativealternativealternativealternativeomnigentagentdockscale-agentexdeer-workflowagentfieldharnessrouter
pairs wellalternativebuilt withpick a node for the why · open it from the panel
Weekly digest
README.md1 min read
The world’s first unified interface for agent harnesses.

Bring the world’s best agent harnesses into your product with one API.

Run Codex, Claude Code, Hermes, DeepSeek Harness, and a growing set of supported harnesses on infrastructure you control. Add or switch harnesses without redesigning your product backend.

GitHub Stars License Docker Pulls UHP Conformance

Discord LinkedIn X

N × M harness integrations → 1 unified interface

HarnessRouter Community Edition gives your product one consistent, OpenAI Responses-compatible interface for running supported agent harnesses. It handles tasks and runs, sessions, streaming, files, artifacts, cancellation, recovery, structured errors, and traces, so your product does not need a separate backend integration for every harness.

One interface. The freedom to choose.

One recorded task, eight Harness × Model configurations. Save 99.8%: cost 0.47–223 credits; latency 1m 25s–4m 36s. Cost and latency vary by task. See the benchmark.

Open source and self-hosted. Community Edition is the Apache 2.0 reference implementation of Unified Harness Protocol (UHP). It packages the Console, Gateway, and Runner in one Docker deployment, with provider keys, state, and files on infrastructure you control.

Quickstart

Start HarnessRouter with one Docker command. On the first launch, HarnessRouter installs the supported harness CLIs you en