Running Codeman at work? Tell me what team and enterprise setups actually need

[Ark0N]

Codeman started as something I built to run my own agents, and most of what I hear about is single-user setups on a home box or a VPS.

But a growing share of repo traffic now arrives from behind corporate proxies and security gateways, which tells me some of you are evaluating this at work, or already running it there. I'd rather ask than guess at what that needs.

If that's you, even a one-line answer is useful:

• What's the setup? Your own machine at work, a shared server, a VM the team hits, containers, something else?
• Who else uses it? Just you, your team, or people outside it?
• What's blocking a wider rollout? Be blunt, that's the useful part.
• What would your security or IT people want before they'd sign off?

What already exists, and I'd like to know whether it's enough

• Multi-user mode (--multiuser): named users, scrypt-hashed passwords, per-user workspaces. Worth being precise about what it is not: it separates workspaces, it does not sandbox users. Every session still runs as the same OS account.
• Docker cases: this is the actual isolation story. One long-lived container per case, workspace bind-mounted at the same absolute path, and credentials seeded into the container rather than shared read-write, so in-container CLIs never write tokens back to the host.
• No telemetry. Nothing phones home. The default bind is loopback only, and reaching it from another device is something you set up deliberately.

What I have deliberately not built, because nobody has asked yet

SSO / OIDC, audit log export, air-gapped or offline install, per-user resource limits, centrally managed policy.

If one of those is the thing standing between you and using this properly, say so and I'll weigh it seriously. I'd rather build the one that's actually blocking people than guess at all five.

And if you looked at Codeman for work and decided against it, I'd genuinely like to hear why. That's more useful to me than silence.