If you work in banking, healthcare, or any other regulated shop, there's a good chance your security team has blocked Claude Code and every other coding agent outright. It phones home to a server you don't control, and that alone fails half the compliance checklists out there.
Coder just closed that gap. Coder Agent Relay launched September 2 as a self-hosted execution environment for cloud coding agents. This week, Claude Code joined the list of agents it supports.
Here's how it works. Each Coder workspace spins up a Claude Code runner that makes an outbound connection to Anthropic's backend. Anthropic still handles billing, account admin, and the model inference.
But the agent's actual work, reading your source code, touching credentials, calling internal services, happens inside a workspace that lives on your own cloud, your own VPC, or on-prem. Your platform team sets the network egress policy. Every workspace inherits it. Every session ties back to a specific user, fully logged.
That's the piece regulated teams have been missing. Anthropic's Cat Wu put it plainly. "The teams that care most about running Claude Code on their own infrastructure tend to be the ones with the most rigorous environmental isolation requirements." The industries that need agentic coding tools the most are exactly the ones that have been locked out of them.
Coder's Josh Epstein backs that up. He says Claude Code is one of the tools enterprise customers ask for by name, and regulated industries have wanted it the most.
One catch. Claude Code support inside Agent Relay is early access only, limited to a handful of design partners right now. Docs are already up at coder.com/anthropic if you want to see the setup before your team gets access.
200 Ways To Make Money With AI
Ready to transform artificial intelligence from a buzzword into your personal revenue generator?
HubSpot’s groundbreaking guide "200+ AI-Powered Income Ideas" is your gateway to financial innovation in the digital age.
Inside you'll discover:
A curated collection of 200+ profitable opportunities spanning content creation, e-commerce, gaming, and emerging digital markets, each vetted for real-world potential
Step-by-step implementation guides designed for beginners, making AI accessible regardless of your technical background
Cutting-edge strategies aligned with current market trends, ensuring your ventures stay ahead of the curve
Download your guide today and unlock a future where artificial intelligence powers your success. Your next income stream is waiting.
Here's why I think this matters beyond one vendor integration. Model providers run the agent loop and bill for it. Customers keep the perimeter, the code, the credentials, the network. That split is what lets a bank's InfoSec team sign off on something they'd never approve if the agent could reach out to arbitrary infrastructure on its own.
I've sat in enough vendor security reviews to know how these conversations usually die. Someone from InfoSec asks where the agent's outbound connections go, the vendor doesn't have a good answer, and the whole thing gets shelved for a quarter (or longer).
If your org has been sitting out the coding agent wave because Legal or InfoSec said no, it's worth another look. Read through the network egress model, or ask Coder about the design partner program, before you write the next memo explaining why your team can't use AI coding tools yet.
And if your platform team already runs Coder for regular dev environments, this bolts onto infrastructure you probably already have. Turning it on looks more like flipping a setting than standing up a new system.
Keep Learning and Building.