# About JetBrains Air

> For the complete documentation index, see [llms.txt](https://www.jetbrains.com/help/air/llms.txt).

JetBrains Air is a system for building software with AI agents. It covers agentic workflows on every level: individual developers, teams, and engineering managers.

JetBrains Air provides one login for all your tools, AI credits as unified currency across vendors, governance and policy controls within your organization, as well as clear visibility into the results, costs, and performance.

* As a developer, you can run your preferred agent approved by your organization either in your favorite JetBrains IDE through [Air in IDEs](air-in-ides-overview.html) or in the CLI through [Air Gateway](cli-about.html).

* As a team lead, you can set up and manage a shared agentic environment in [Air Teams](what-is-air-teams.html).

* As an engineering manager, you control the rules and permissions for agent-driven development for the whole organization through [Air Governance](air-governance.html).

* At any level, you can help coding agents understand any codebase faster and cheaper with [Air Context](air-context.html).

Different parts of JetBrains Air are deeply interconnected, meaning that you can start an agent session locally in your IDE and then hand it off to your team's remote agentic environment. Or if some automation finished not quite the way you wanted, you can pull it into your IDE and continue the interaction with the agent there.

There are policies that govern the rules and permissions at every level: organization, team, and for every individual member of your organization. And not only human members: service accounts can also have policies specific to them. And these rules and permissions propagate across the whole system.

All parts of JetBrains Air share the same login credentials and consume credits from a shared pool regardless of what agent is running. The same credits pay for cloud compute where the remote session is running, for context indexing and storage.