The Pulse
JetBrains Turns Air Into a Multi-Vendor Agentic Development System
JetBrains has expanded Air from an agentic development environment into a product system spanning developers, teams, and organizational governance. The company says Air will connect JetBrains IDEs, autonomous agents, shared context, workflo
AI.info Team ·
“The goal is not more code. It is software that developers, teams, and organizations can understand, verify, and stand behind.”
Kirill Skrygan, author of the JetBrains announcement
JetBrains is turning Air into a broader system for agentic software development, linking coding agents, JetBrains IDEs, team workflows, and organization-level controls in one product family. The company announced the expansion on September 22, describing Air as an “open, coherent system of products” that operates inside and beyond its IDEs.
The move takes JetBrains beyond its traditional focus on the individual developer workbench. The company says it has spent 26 years building developer tools and is now targeting the wider process through which agentic work is initiated, executed, coordinated, reviewed, and governed.
Air Connects Developers, Teams, and Governance
JetBrains Air will combine products aimed at different levels of software delivery. Air in JetBrains IDEs is intended to let developers direct and orchestrate agents while inspecting their output with JetBrains’ code intelligence. Air Teams will coordinate workflows involving developers and autonomous agents. Air Governance, formerly called JetBrains Central, will handle organizational policy, visibility, auditability, cost management, and accountability.
Junie, JetBrains’ coding agent for professional software development, will be supported across all Air surfaces. The company also says Air will remain open to agents and services that JetBrains does not build, reflecting its view that developers will use different models and agents for different tasks.
That openness depends in part on the Agent Client Protocol, or ACP. JetBrains says the protocol standardizes the connection between an IDE and an agent’s full harness, including planning, tools, model routing, and observability. Through the ACP Registry, developers will be able to discover and run compatible agents while continuing to work inside JetBrains IDEs.
JetBrains Targets the Cost of Fragmented Agent Work
The announcement frames Air as a response to a gap between individual adoption of coding agents and the infrastructure organizations need to manage them. JetBrains says agents are already producing practical value for many developers, but the organizational costs appear in review, rework, security, infrastructure, and AI spending.
The company identifies a series of questions that become harder as agent use spreads: which agents can access company code, where data can go, which outputs need human review, what happened while an agent worked remotely, who approved a change, and how that change was verified.
Air Governance is designed to give organizations one view across providers and tools. JetBrains says the system will preserve context, enforce policies, attribute spending, and record what happened even when different vendors produce different parts of a software change.
Air’s Multi-Vendor Strategy Puts Control Above the Model
JetBrains is not positioning Air as a replacement for every other coding agent. The company says it will build and support its own agent, Junie, but will not require customers to use it. Instead, Air is meant to provide a common place to run, observe, govern, and account for development work across models, agents, and services.
JetBrains argues that model rankings and capabilities change often enough to make a single-vendor commitment difficult to justify. Teams within the same company may also choose different tools for different jobs. The challenge, in the company’s telling, is that each additional service can fragment context, spending data, policy enforcement, and records of agent activity.
Supporting multiple agents is only the starting point, JetBrains says. The layer above those tools is intended to provide shared context, unified policies, a single cost view, and a record of how a change was produced.
JetBrains Brings Code Intelligence Into Agent Workflows
JetBrains’ technical argument rests on its existing code intelligence. The company says agents can spend time and money rediscovering information already contained in a codebase, while deterministic analysis can help them retrieve that information directly.
That approach is intended to make agent work easier to understand and verify across different models. JetBrains says richer context will eventually draw from code, architecture, repositories, runtime behavior, and organizational knowledge, allowing work to move between developers, models, agents, and services with less loss of context.
The company also expects more work to begin outside the traditional editor session. Future Air workflows may be triggered by repository events, schedules, and delivery processes rather than by a developer opening an editor and entering a prompt.
JetBrains Plans Rolling Air Releases
JetBrains says Air will arrive through a rolling series of releases, with separate disclosures for features available now, features entering preview, and longer-term plans. The company says future versions will extend into mobile and remote experiences so users can initiate, monitor, review, and continue agentic work across different environments.
JetBrains has not announced a complete product timetable for Air Teams, Air Governance, or the IDE experience. Its September 22 announcement instead establishes the structure: an open system built around agents, code intelligence, team coordination, and organizational responsibility.