Koda updated koda-cli 0.3.1: the assistant can now be integrated into a development environment

Koda released koda-cli 0.3.1 with improved Agent Client Protocol support. The main idea of the update is to move work with the AI assistant from the terminal to the interface of a compatible development environment, where the developer already spends most of their day.

What was updated

In the new version, the Koda team improved ACP support and showed how to connect the assistant to popular IDEs. This is not a complete abandonment of the command line: for initial installation of koda-cli, the terminal is still needed. But after that, the scenario changes.

The assistant can be launched in the background from the development environment and you can communicate with it through the built-in interface without switching between windows every time you need to ask about code, fix a file, or go through a multi-step agent workflow. For koda-cli, this is an important shift in positioning. Previously, the word CLI itself implied that the main experience would be tied to the terminal.

Now Koda is showing a different route: CLI remains the engine and installation point, while daily work moves to the client GUI. This approach makes the tool easier to understand for teams where developers, analysts, and tech leads want to use the agent next to the code rather than in a separate console.

How ACP works

Agent Client Protocol is structured like a standard client-server scheme. The assistant runs in the background, and the development environment acts as a client and communicates with it via HTTP through WebSocket or JSON-RPC. In practice, this means that the chat, agent actions, and the entire task execution process are displayed inside the IDE, while the assistant itself continues to work as a separate service. For the user, this is closer to native integration than running a script from the terminal.

The entire agent workflow is reflected in the IDE interface, not in

the terminal.

Koda separately emphasizes that this principle is not tied to a single application. If your favorite editor or IDE already supports ACP, koda-cli can be used there too. If support is missing or bugs are found in the integration, the team asks to report issues through issues. This is an important signal for the market: around AI assistants, not only a set of models is gradually forming, but also a layer of compatible protocols that reduces dependence on a single vendor and single interface.

What changes in workflow

For everyday development, such an update removes one of the most annoying details — constant jumping between code and the terminal. When the assistant interface lives in the same environment where the project is open, it's easier to maintain context, read responses next to files, and quickly check the result. This is especially noticeable in tasks where the agent does not simply respond with text, but sequentially executes several steps and should show progress as it works.

• Code discussion and edits within the editor, without a separate console
• Running agent scenarios in the background from a compatible ACP client
• Using the same assistant in different software if it understands the protocol
• More intuitive onboarding for teams uncomfortable with the terminal interface

At the same time, the update does not cancel the basic role of CLI. Installation and initial setup still need to go through the command line, and then you can connect a client with ACP support. That is, Koda does not abandon its original architecture, but expands it. This is a pragmatic path: maintain flexibility of CLI for experienced users and at the same time give a more familiar entry point for those who want to work with the agent from the IDE window.

What this means

Koda is betting not on a closed interface, but on compatibility through protocol. If ACP continues to attract new clients, koda-cli can exist in several development environments simultaneously without rewriting the logic for each one. For the AI tools market, this is a good sign: competition is shifting from a fancy shell to the quality of the assistant, integration speed, and how convenient it is to embed it in real workflow.