Repository navigation
One AgentZ, every team, one admin #109
HighnessAtharva
announced in
Announcements
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
One AgentZ, every team, one admin
Most AI tools stop at one person and one API key. AgentZ starts at the company.
Sign up and you get an organization. Inside it you create a workspace per team. Security, DevOps, HR, sales. Each workspace stays walled off from the others.
You configure the shared parts once. Skills, sandboxes, MCP connections, inference providers. Every new workspace inherits them, so nobody rebuilds the same setup five times.
Access is the product surface
Roles decide who sees what. A role is a named set of capabilities. Attach it to a person, or to a team.
Sharing is an authorization check, not a link. You can only grant capabilities the target already holds in that workspace.
Onboarding follows the same rules
Turn on social sign-up and set the roles and teams that new members land in. No manual cleanup after the fact.
How does your team split? Tell us below.
All reactions