harnessrouter

One API that runs agent harnesses like Codex and Claude Code as backends

Documentation ↗ Website ↗ Repository on GitHub ↗

Stars: 4.2k
License: Apache-2.0
Last commit: Nov 2026
Last release: Nov 2026
Language: Python

Missed the latest GitHub check; these facts may be older.

Overview

HarnessRouter wraps existing agent harnesses (31 listed, including Codex, Claude Code, OpenCode, Aider, OpenHands and goose) behind an OpenAI Responses-compatible API. The harness is chosen per request with metadata.harness_id. It handles persistent sessions, SSE streaming, files, cancellation, idempotency keys and W3C trace propagation. Community Edition runs as one Docker container with a Console, Gateway and Runner, using your own provider keys.

Who it is for: Developers embedding coding-agent harnesses as backends in their own products

Strengths

OpenAI Responses-compatible API; harness switched per request via metadata.harness_id
31 harness bases, with a support matrix per provider
Idempotency-Key retries, W3C traceparent propagation and generated OpenAPI spec
One container with a persistent /data volume; Console analytics disabled in CE

Weaknesses

Harness is fixed once a setup is saved; switching needs a new setup
Container must run as root to manage per-session users; --user is unsupported
Starter Kits are licensed under different terms than Community Edition
Managed scaling and serverless sandboxes exist only in the hosted Cloud

What it needs

no GPU
Docker + Compose
Needs Docker, Model provider API key
Models: OpenAI Responses API, Provider API keys (bring your own)
port 3000

Also in Agent platforms

See all 92

Also in Agent platforms
Rank	Project	Score
1	AutoGPTRole-based AI agents coordinated by one lead agent, plus a visual workflow builder Live demo ↗ (opens in a new tab)188k starscustom license	95 out of 100
2	daily_stock_analysisLLM-driven daily analysis of watchlist stocks across six markets, pushed to chat apps71k stars+4,757 (+7.2%) in 30 daysMIT	95 out of 100
3	LangflowVisual builder for AI agents and workflows, exposed as API or MCP server156.6k starsMIT	94 out of 100
#4	deer-flowAgent harness running sub-agents, memory and sandboxes with pluggable skills85.3k stars+1,729 (+2.1%) in 30 daysMIT	93 out of 100
#5	DifyVisual builder for LLM workflows, RAG pipelines and agents, with APIs Live demo ↗ (opens in a new tab)161k starscustom license	92 out of 100

How to choose

Decide between a visual builder (faster to start) and code-first (easier to test and version).
Check how workflows are exported; vendor-specific JSON makes migration costly.
Look at the license for the server part; several are AGPL or source-available.

Health

Releases: Not checked
Top contributor: Not checked
Security advisories: Not checked
Package advisories: Not checked
OpenSSF Scorecard: Not checked

Not read for this project yet: each refresh reads the facts of a few projects. Bots and coding-agent accounts are left out of the commit counts. Not part of the score: how we check.

License in plain words

Apache-2.0 Permissive

Use it at work: Yes
Change it and rebrand it: Yes
Sell it or host it for others: Yes

Your part: Keep the notices and mark the files you change. It also grants you the contributors’ patent rights.

A plain summary, not legal advice. Read the license on GitHub ↗ (opens in a new tab)

Set it up

Needs before it works: Docker, Model provider API key.

Bring in an engineer who knows harnessrouter. By the hour for upkeep and fixes, or a fixed price for a build. Our own engineering service; it never changes a rank.

Where to run it

RackNerd ↗ (opens in a new tab)Budget KVM VPS from $26.99 a year to run Docker apps yourself.
RepoCloud ↗ (opens in a new tab)One-click hosting for about 890 open-source apps, each on its own VPS.

Referral links: if you sign up through one, we get a commission. They never change a rank.

Ask your AI assistant

The rankings as an MCP server for Claude, Cursor, ChatGPT and any MCP client. Free, no key, read-only.

Setup for other apps

Claude Codeclaude mcp add archestack -- npx -y @archestack/mcp

Connector URLhttps://mcp.archestack.com/mcp

Then ask: Compare harnessrouter with its top alternatives on Archestack

#77 of 92 in Agent platforms on Archestack, as of 12 Nov 2026.

Show it in your README or on your site

The badge below goes in a README; a card with the rank, score and signals goes on a site or blog. Both update twice a day.

Show code

README (Markdown)

[![Archestack](https://archestack.com/badge/harnessrouter/harnessrouter.svg)](https://archestack.com/best-of-ai/projects/harnessrouter/)

Site or blog (HTML)

<iframe src="https://archestack.com/embed/best-of-ai/harnessrouter/" title="harnessrouter on Archestack" width="400" height="270" style="border:0;max-width:100%" loading="lazy"></iframe>

How badges work Preview the card

Best of Open-Source AI on GitHub

Each review comes from the project's README and facts we check on GitHub. The fix form opens with harnessrouter filled in.

Chat and assistants

Agents and MCP

RAG, search and memory

Voice

Image and video

Workflows

Coding with AI

Apps and SaaS

Models and ML

Security

Learning

By field