We've been thinking about building an AX Studio which makes it easier to create a local cluster and test things locally. Once tested locally, agentic tasks can be run against a production cluster without changes.

👌🙏

open source it :)

Is it open source?

Very important imo, easier to grasp the concepts. Nice!

It will be filled with 20 more buttons and 100 more sub-links in the next 6 months. Button_1: agentic_task GPT 5.1 sub_link1: deeper deep sub_link2: decision mode Button_2: Gemma_task Argon 1.1.1.1 sub_link1: deepest deep sub_link2:

Looks good! Perhaps adding a tool similar to k9s would make things even more convenient.

Make sure it has an MCP.

The “without changes” part is the hard bit. My unattended job passed locally, then died overnight on a Windows console encoding error from non-ASCII log text. I’d want the studio to match prod locale and encoding, not just the cluster.

Stuff like this is also useful for people curious to learn about the stack. GUIs do spatial demonstrations better than cli man pages

local to prod with no changes is the right goal. What usually breaks is config and permissions, not the code. If the local cluster runs with the same service accounts and policies as prod, just scoped down, the agent can't pass locally and then hit a permission wall in prod

local tests miss prod permissions and real data, so the first prod run still needs a human ok

Would still have an agent doing the configuring, but having a way to visualize what’s being set up, as it’s set up, helps with my understanding.

That would be so useful! Would the agents be able to select which parts of prod to run locally to save resources? Will it be easy to capture/define service dependencies?

local-to-prod parity is where this gets hard. configs drift, quotas only kick in at real traffic. what's the local version of a rate limit you only hit in prod?