Show HN: First background agents in Jetbrains IDEs [video]

7 kevo1ution 1 8/26/2025, 4:50:04 PM youtube.com ↗
TLDR: made the first background coding agent that has an isolated workspace and runs locally

Howdy - I’m Kevin, co-founder of Firebender, and we built the first background coding agent in android studio / Jetbrains!

Why not just use Cursor background agents or OpenAI Codex?

Both of these require setting up a cloud container and cloning your existing developer environment, and maintaining it. Then when you want to iterate on changes as AI inevitably makes a mistake, you either throw away the work, or have to pull down the branch and clean it up. This feels really clunky. With Firebender, background agents run locally in a lightweight git worktree/IDE tab. This means when the agent is done, you can easily clean up the changes and run code with a few clicks.

Under the hood, the agent behaves similarly to claude code (didn’t want to reinvent the wheel), but also leverages all of the hooks into IntelliJ sdk like go-to-definition, find usages, auto-imports for accuracy, and it gives a cleaner visual UI for reviewing changes and merging them. You can use any frontier model like gpt-5/sonnet-4 as the base.

We’ve had to do quite a bit of reverse engineering of the IntelliJ codebase to cleanly set up and manage the isolated environment, and I think you’ll appreciate the simple UX of hitting cmd+enter to run a background agent anywhere.

get started docs: https://docs.firebender.com/get-started/background-agents

download the plugin: https://firebender.com

Would love to get your feedback to help us improve the tool for you! Thanks!

Comments (1)

sarthaksrinivas · 59m ago
Curious about your reverse engineering efforts - hat was the trickiest part about getting Firebender’s background agents to run in fully isolated workspaces in Jetbrains IDEs? Did you run into issues with IntelliJ’s project management / sandboxes behind the scenes, especially when trying to keep the agent’s changes cleanly separated from your main code? How did you manage those edge cases or unexpected side effects

Really interesting launch, there’s lots like this to do in order to improve devex in the Jetbrains ecosystem