Many people still aren't using AI for coding. I talk to a dozen companies every week, and this is a recurring topic: • They don't trust model providers • They don't trust agents (models sometimes do crazy things) • It's hard to standardize agentic tools across large teams AI development is no longer just about models assisting people; we now have agents writing code by themselves! Easy to implement on your laptop, but very challenging to do at a company level with hundreds of developers, complex rules, and strict privacy requirements. For anyone working as part of a team, here’s a great solution: Check out @coderhq. It's open-source, runs anywhere (cloud, on-prem, or air-gapped), and connects to virtually any IDE out there. They collaborated with me on this post. Coder lets you create and manage secure environments that you can share across your entire team. Your developers can use this environment to write code, or you can have Claude Code (or any other Agentic coding platform) use the same environment as well. • No local configurations • No security or compliance headaches. • No dependency on what OS or hardware your developers use. For example, I can have my team work in an environment configured for Cursor, and another environment with GPU access dedicated to Claude Code. Best part: • You can audit every action inside an environment • You can standardize how environments are configured across teams It's pretty awesome, and remember: this is all open-source! Here is a link: fandf.co/4nqaDK4