Show HN: OpenRig – a control plane for multi-agent coding topologies

[mschwarz]

Hi HN, I’m Mike, the founder of OpenRig.

I built this because my Claude Code + Codex setup kept forming little "topologies" of long-lived agents that worked well together, but the terminal sprawl was intense. So I built a primitive the agents could intuitively reach for to save and recreate these setups on the fly. This then led to more agent-first primitives like coordination, declarative workflow patterns, workspaces, etc.

Several months in and these "rigs" I manage with openrig require a lot less babysitting and I can manage more projects at once without getting overwhelmed.

The short version: OpenRig is a way to save and operate that topology instead of rebuilding it by hand every time.

Demo video: https://youtu.be/yCFSRnPDFqY?si=n5e627d0CU3X3bmE

GitHub: https://github.com/mvschwarz/openrig

[jonnyasmar]

Super interesting mental model here. I'm building a tool, called atrium, that set out to solve similar problems, but I landed on a pretty different model -- workspace > room > stack > pane. I'm really curious to understand more about your concept of topologies, though. How exactly does it manifest in practice? Are they essentially organized units of collaborative work that your agents are executing? Has this unlocked different ways to solve problems for you than you were doing before or is it primarily organizational sugar?

At any rate -- really cool concept. Wish you the best of luck with it!

[mschwarz]

Yes exactly, topologies, aka rigs are basically ways of organizing work that benefits from agent collaboration, but it unlocks many benefits. One of the main benefits is "Context Domains", described in the linked video. Where you can extend the context window for the task you are working on by distributing it across multiple agents. Each agent or pod of agents works on a slice of the task together, sharing state, staying in their lane of expertise. The rig becomes a sum of its parts. A 10 agent rig, each agent with a 1 million context window, could effectively be viewed or operated as if it had a 10 million token context window. For my day job I work on very large brownfield codebases and this was one of my early solutions to being able to work on broad cross-cutting features. I don't rely on this alone but this general ability is quite useful for lots of things.

[NBenkovich]

Thanks for sharing! I like the analogy with climbers! Is OpenRig compatible with the A2A protocol?

[mschwarz]

Thanks! It doesn't use A2A but that might be a good roadmap item, there's a ton of overlap with it. OpenRig already has a lot of A2A like wiring built-in but simpler and intentionally stripped down to keep things inspectable by a single human operator trying to manage a fleet of agents. It just uses TMUX for agent-to-agent messaging- super simple but effective. For tasks, handoffs, etc there is a custom event stream and queue system.

[NBenkovich]

Got it! But what do you think about isolation of agents? If you use tmux, each agent has the same network, filesystem and etc

[mschwarz]

Yeah so right now OpenRig isn't trying to solve that problem yet, but thats a very interesting direction to take it someday. It currently assumes you the operator are running coding agents on a host you control, like your mac or a VPS, and you actually want all the agents to have access to the same network and file system, so they can work together without friction. It basically takes the core single-user UX of CLI coding agents like claude code and asks the question "what if i could scale this up and run 10, 50, 100 of these things in parallel." OpenRig is the result of the last 9 months of exploring that branch of the AI tree very heavily.

[NBenkovich]

Got it! Thanks for answering!

[mschwarz]

I posted a much earlier preview a few weeks ago. This version is the first one I’m actually inviting other agent-heavy developers to try, and I’d love feedback on where the mental model is confusing.