Dev Tools
mission-control
Self-hosted control plane for AI agents: dispatch tasks, review runs, track spend, and operate OpenClaw, Claude Code, Codex, and other runtimes.
★ 6.3k starsMIT
View guide →One API for every agent harness — run Codex, Claude Code and more through a single self-hosted interface
harnessrouter is an open-source dev tools project by HarnessRouter (HarnessRouter/harnessrouter on GitHub). It has 2,929 GitHub stars and 299 forks, is released under the Apache-2.0 license, and was last updated on GitHub on 8 October 2026.
Decide which harnesses you want behind one API and which model providers you will connect with your own keys.
Pull and run the official Docker image with a persistent data volume.
Wait for first launch to install the enabled harness CLIs, sign in to the console, and add a model provider under Bring Your Own Key.
Submit a task through the console or the OpenAI Responses-compatible API and watch it stream, then try switching harnesses.
Implements UHP, an open standard, so harnesses become plug-and-play agent backends behind one integration.
Run Codex, Claude Code, Hermes, PI, DSH and more, and switch harnesses per task to optimize cost and latency.
Provides an OpenAI Responses-compatible API for product integration.
Persistent sessions, streaming progress, files and artifacts are handled by the router.
Built-in cancellation and structured failure handling across harnesses.
Community Edition runs on your infrastructure with your own provider keys; data persists in a Docker volume.
Clone HarnessRouter (the quickest run path is the Docker image in step 2).
git clone https://github.com/HarnessRouter/harnessrouter.git
cd harnessrouterStart the self-hosted Community Edition exactly as the README quickstart shows.
docker run -d --name harnessrouter \
-p 127.0.0.1:3000:3000 \
-v harnessrouter:/data \
harnessrouter/harnessrouterWait for '[harnessrouter] ready on :3000' in the logs, then open http://localhost:3000 and sign in (default harnessrouter / harnessrouter).
docker logs -f harnessrouterHarnessRouter (the team behind harnessrouter.ai) maintains it; this repo is the open Community Edition of their unified harness interface.
The Community Edition is free and Apache-2.0 licensed, self-hosted on your infrastructure. Harness CLIs installed on first launch keep their own upstream licenses.
Yes — you bring your own model provider keys (added in the console under Bring Your Own Key); model usage is billed by your provider. You do not need a HarnessRouter API key to run console tasks.
Direct integration means handling sessions, streaming, files and failures separately for every harness. HarnessRouter gives one API and lets you switch harnesses freely.
Run the Docker command from the README, wait for the ready log line, open http://localhost:3000, add a provider key, and run a task.
Self-hosted control plane for AI agents: dispatch tasks, review runs, track spend, and operate OpenClaw, Claude Code, Codex, and other runtimes.
Self-hosted dev tools: an intelligent agent that understands your code and cra.
A local-first Agent OS — drive Claude Code, Codex and OpenCode from your browser.
A kanban board for coding agents — plan tasks, run Claude Code or Codex in parallel, review diffs
AI coding agent skills for KiCad electronics design. Works with Claude Code and OpenAI Codex. Analyze schematics, review PCB layouts, EMC pre-compliance, SPI…
Cross-platform desktop all-in-one manager for Claude Code, Codex, OpenCode and other coding agents
The most-starred other Dev Tools projects in the LibreAI catalogue — the realistic shortlist if harnessrouter is not the right fit.
The open source coding agent
The model-definition framework behind modern AI — thousands of pretrained transformers for text, vision, and audio.
Anthropic's agentic coding tool that lives in your terminal and works across your whole codebase
Cross-platform desktop all-in-one manager for Claude Code, Codex, OpenCode and other coding agents
OpenAI's lightweight coding agent that runs locally in your terminal
Free and open — grab the source, star the repo, and start creating.