AI tools for Kotlin development

AI Assistant

AI Assistant provides AI-powered assistance directly in JetBrains IDEs, such as in IntelliJ IDEA, as well as in Android Studio. You can use it for interactive development tasks where you want to stay in control of each change.

The AI assistant provides:

• Access to AI agents, including Junie, Claude Code, OpenAI Codex, and any third-party agents that support the Agent Client Protocol.

• Context-aware AI chat using cloud-hosted models, such as Gemini, GPT, and Claude, as well as your own local models.

• AI-assisted code completion and next edit suggestions.

Learn more about AI assistant integration with JetBrains IDEs.

Agent Client Protocol

The Agent Client Protocol (ACP) is an open protocol for connecting AI agents to IDEs and code editors. ACP defines a common protocol for AI agents and development tools to communicate without separate integrations for each agent and editor combination.

JetBrains IDEs support ACP, allowing you to use compatible AI agents within your IDE. You can choose between different AI agents while working with Kotlin-aware IDE features such as navigation, inspections, refactoring, debugging, and project analysis.

The ACP registry provides access to multiple agents, including Claude Agent, Cursor, GitHub Copilot, OpenCode, and others. See the full list of supported agents in the ACP registry.

Junie

Junie is a JetBrains AI agent. You can use Junie in JetBrains IDEs and Android Studio, from your terminal, or in headless mode in CI/CD pipelines. You can also integrate Junie into your GitHub workflows.

Junie is designed for tasks that require more than a single code suggestion or chat response. Use Junie for development tasks that involve multiple files or require planning and execution. You can ask it to implement a feature, update code across multiple files, add tests, or perform maintenance work.

When Junie runs in an IDE, it can also use IDE capabilities such as project indexing, code navigation, inspections, refactorings, debugging, and framework-aware project analysis.

Learn more about Junie.

Third-party AI agents

Many third-party AI development tools support Kotlin. They are available as IDE extensions, standalone editors, command-line tools, and cloud-based development environments. For example:

• GitHub Copilot

• Google Gemini

• Claude Code

• OpenAI Codex

Choose a third-party tool if it matches your preferred development environment or offers capabilities that fit your workflow. Many of these tools support Kotlin code generation, explanations, test creation, and refactoring.

You can use third-party tools independently or connect compatible agents to JetBrains IDEs through ACP.

MCP servers

The Model Context Protocol (MCP) connects AI models to external data sources, tools, and systems. JetBrains maintains several MCP servers that can make your Kotlin experience more productive:

• The JetBrains IDE MCP server exposes IDE capabilities. Using the server, an AI agent can use IDE features such as project indexing, code navigation, refactoring, inspections, and build execution. This gives the agent a better understanding of your Kotlin project as well as more efficient means of generating and evaluating code.

• The MCP Kotlin SDK is a Kotlin Multiplatform implementation. It helps you build AI-powered applications in Kotlin and integrate with LLM surfaces across the JVM, WebAssembly, and iOS.

• For Kotlin Multiplatform projects, the klibs.io MCP server helps agents access the catalog of available multiplatform libraries to more efficiently look for existing solutions.

• For Compose Multiplatform projects, the Compose Hot Reload MCP server allows agents to directly interact with the reloadable app: trigger reloads, take screenshots, read semantic trees, and so on.

Kotlin AI skills

Kotlin AI skills are reusable instructions that guide an AI agent through Kotlin development tasks. They help agents perform these tasks more consistently.

Use Kotlin AI skills when you want to guide an agent toward idiomatic Kotlin patterns, Kotlin coding conventions, and project-specific expectations. Skills help AI agents perform tasks such as writing Kotlin code, explaining language features, generating documentation, creating tests, reviewing code, or applying migration guidance.

Kotlin AI skills can be used with different agents and workflows, including IDE-based agents, command-line agents, and external AI tools that support reusable instructions.

Learn more about Kotlin AI skills.

Kotlin-specific acceptance criteria

Kotlin Multiplatform projects in particular are complex enough for agents to lose track of the entire project structure and consequences of specific changes.

To help the agent, you can include the following examples as your general (AGENTS.md) or task-specific success criteria:

• Run target-specific tests after introducing changes whenever such tests are available.

• Verify that all configured KMP targets successfully build before considering a task complete.

• Review the implementation for platform-specific APIs leaking into common code to avoid agents (or developers) accidentally using these APIs in common code later on.

JetBrains Air

JetBrains Air is the Agentic Development Environment (ADE) for engineering teams building products with AI. With Air, you can provide context for each task, choose the agent, model, and execution environment, and then review or refine the resulting changes before applying them to your code.

Use Air when you want to delegate a defined coding task to an AI agent, isolate AI-generated changes from your local working copy, run multiple implementation tasks in parallel, or turn recurring development tasks into scheduled or event-driven automations. You can run tasks in your local workspace, isolated Git worktree or Docker container, and JetBrains-managed cloud environment.

You can work with JetBrains Air through:

• Air desktop app – Run local and cloud tasks from a desktop application.

• Air on the web – Run, monitor, and manage cloud tasks and automations from a web browser.

• AI Assistant in IntelliJ-based IDEs – Start cloud tasks and review results without leaving your IDE. Work with the same tasks in the Air desktop app or in the web version.

Learn more about JetBrains Air.

JetBrains Central

JetBrains Central is a platform for agentic software development across an organization. It connects AI agents, development tools, and infrastructure so that agent-driven work can run, be monitored, and be managed across teams, with visibility into results, costs, and performance.

JetBrains Central Console is the web interface for organization-level AI governance in JetBrains Central. Organization administrators can use the Console to manage access and policies, monitor AI usage and spending, analyze adoption, and control which AI models and features teams can use.

Learn more about agentic software development.