Air Context

Air Context uses semantic search to solve the problem of code search and code exploration for agentic tasks. It indexes your organization's codebases and lets AI agents query them semantically instead of performing ad-hoc searches and file reads. In JetBrains Central Console, admins control which repositories are indexed and how.

Most problem-solving and feature development tasks require agents to perform extensive code exploration to understand the current state or get a reference for the planned changes. In practice this means running searches, spinning up exploration subagents, and reading files – activities that often consume a lot of time and tokens. Air Context replaces this exploration with semantic search over a prepared index.

Specifically, Air Context provides the following features:

Incrementally indexes code, docs, manifests, and other artifacts at scale, using a hybrid approach that combines full-text search and vector embeddings.
Exposes semantic search tools and skills that agents call to “ask” for relevant files, APIs, tests, patterns, or examples instead of manual grep-style exploration.
Supports multi-repo awareness, so agents can search beyond the current project for uses such as finding dependencies, assessing impact radius, or reusing code across repos.
Integrates with JetBrains Central Console to provide semantic search for all supported AI-powered tools and features.
Lets agents understand and navigate large, complex codebases faster and cheaper, while maintaining or improving solution quality.

Indexing

Air Context distinguishes between Indexing by JetBrains Central and On-demand indexing. All indexed repositories fall into one or both of the two categories, depending on where the request to index them originates from, and the list of repositories on the Air Context page has a tab for each category:

Indexing by JetBrains Central: An admin connects JetBrains Central Console to GitHub and chooses which of the organization's repositories to index. JetBrains Central then indexes them on a schedule and stores the indexes. Repositories can also come from a self-hosted version control system connected on the Hosted VCS page. The key difference is that the code of these repositories is processed by JetBrains.
The same applies to repositories that users connect in Air Teams: JetBrains Central indexes them automatically.
On-demand indexing is triggered by running the Air Context CLI (jbcontext) manually or in CI pipelines. It’s permissive by default and allows an organization’s employees to index anything until an admin explicitly blocks it in JetBrains Central Console. The repository code is processed on the organization's own infrastructure, and JetBrains doesn't access the repositories directly. For what is sent to the Air Context server during indexing, see Data flow.

GitHub is not connected by default, so initially the Indexing by JetBrains Central tab lists only the repositories that have already been indexed, for example, those that users connected in Air Teams. Connecting to GitHub is required to discover the rest of the organization's repositories and to start indexing them. The same repository can be indexed both by JetBrains Central and on demand; the two indexes are independent of each other, so such a repository appears on both tabs, with its own status on each. For details on what data Air Context stores about indexed repositories, see Data privacy.

The Air Context page

The Air Context page in JetBrains Central Console is where you manage the indexing settings of your organization and the indexed repositories. The page has three main areas: the GitHub pane, the Settings pane, and the list of repositories.

GitHub connection

The GitHub pane at the top of the page shows the state of the GitHub connection. Connecting is required to discover the organization's GitHub repositories and to start indexing them by JetBrains Central. For details, see Connect to GitHub.

Settings

The Settings pane holds the organization-wide indexing settings.

Context indexing: Controls whether repositories in this organization can be indexed at all. Blocking it stops both indexing by JetBrains Central and on-demand indexing.
On-demand indexing: Controls whether repositories can be indexed on demand through the Air Context CLI. This setting is nested under Context indexing.

Each row has a button to block indexing or allow it again. While indexing is blocked, a message below the pane lets you delete the existing indexes.

List of repositories

The list of repositories is split into two tabs: Indexing by JetBrains Central and On-demand indexing.

Each tab has the following columns:

Column	Description
Repository	The repository name or path.
Status	The current indexing state of the repository, such as Indexed or Not indexed. For the full list of statuses, see Manage repository indexing.
Scheduled at	When the next periodic indexing run of the repository is scheduled. For repositories indexed on demand, the column shows a dash.
Last indexed	When the repository was last indexed. The column can also show remarks such as No periodic indexing, Run on demand, or Last successful index.

You can search the list, and use the controls in each row to manage indexing for individual repositories.

20 October 2026