# CC GUI: A Claude Code and Codex GUI for IntelliJ IDEA

> CC GUI wraps terminal AI coding CLIs in a JetBrains plugin, with Claude Code and Codex as the headline engines and six more marked Beta. It is a convenience layer over tools you still have to install yourself.

**zhukunpenglinyutong/jetbrains-cc-gui** — Jetbrains Claude Code and Codex GUI Plugin

- Repository: https://github.com/zhukunpenglinyutong/jetbrains-cc-gui
- Website: https://plugins.jetbrains.com/plugin/29342-cc-gui-claude-or-codex-
- Stars: 6,740 · Forks: 665
- Language: Java
- License: MIT
- Published: 2026-09-22 · Updated: 2026-09-22 · Language: en
- Canonical page: https://hysenlabs.com/projects/zhukunpenglinyutong-jetbrains-cc-gui

## What CC GUI actually replaces

Claude Code, OpenAI Codex and the other supported engines are command-line tools. You run them in a terminal, read diffs as text, and manage conversation state by scrolling. CC GUI puts a Swing-based IntelliJ IDEA plugin in front of that workflow: a chat panel, @file references, image attachments, a code DIFF view, file navigation, session history with search and favorites, and message export. The README describes the target as "a visual interface for Claude Code, OpenAI Codex and more AI coding CLIs, making AI-assisted programming more efficient and intuitive." The audience is JetBrains users who want the terminal agent loop without leaving the IDE, and who are willing to install the underlying CLIs separately. It is not a model. It does not provide API keys, and it does not run inference on its own.

## How the plugin, the webview and ai-bridge fit together

The repository layout tells you more about the architecture than the README does. There are three moving parts: src/ holds the Java plugin code, webview/ holds a frontend with its own package.json, and ai-bridge/ holds a Node.js layer with its own package.json. The Dockerfile confirms the split, copying webview/package.json and ai-bridge/package.json first so npm dependencies cache independently of source changes, then running the Gradle task buildPlugin, which the Dockerfile comment says transitively runs buildWebview and packageAiBridge. So the chat interface is web technology rendered inside the IDE, and ai-bridge is the process that talks to the external CLIs. The Dockerfile also documents a practical detail: it uses npm install rather than npm ci because "the upstream lockfiles drift from package.json" and npm ci aborts on a missing transitive entry. That is an honest admission of dependency drift, and it means reproducible installs are not guaranteed from the lockfiles alone. The packaged plugin is roughly 40MB according to the build comment, which is the cost of bundling both a webview and a Node layer.

## Installing CC GUI in IntelliJ IDEA

The README points installation at the JetBrains Marketplace listing rather than describing manual steps, so the normal path is the IDE's own plugin manager. The marketplace page is the source of truth for which IDE builds are supported. If you want to build the plugin yourself instead of installing the published artifact, the README gives the local development sequence. First install the frontend dependencies, then the bridge dependencies, then run the plugin in a sandbox IDE. Each command runs from the repository root and the README shows them as separate steps.

```bash
cd webview
npm install
```

```bash
cd ai-bridge
npm install
```

```bash
./gradlew clean runIde
```

The runIde task launches a separate IDE instance with the plugin loaded, so your normal installation is untouched. For a distributable artifact, the README gives buildPlugin and notes the output location and size.

```sh
./gradlew clean buildPlugin
```

After that command the README states the generated plugin package is in build/distributions/ and is approximately 40MB. If you would rather not install a JDK and Node toolchain locally, the repository ships a Dockerfile that builds the same zip. The Dockerfile header gives the three commands: build the image, then run it with a host directory mounted at /out, and optionally override the target IDE with the TARGET_IDE environment variable, where IC is IntelliJ Community and the other documented values are PC, PY and RD.

## Engine support is uneven, and the README says so

Eight engines are listed. Only Claude Code and OpenAI Codex appear without a qualifier. Grok CLI, Kimi CLI, OpenCode, PI CLI, OMP CLI and DeepSeek Harness are each labeled Beta in the README. That is six of eight. Treat the Beta label as the project's own statement about stability, not as a formality: an engine marked Beta may change behaviour between releases, and the README does not publish per-engine compatibility notes or a support matrix. A second limitation is structural. The plugin drives external CLIs, so a broken or unauthenticated CLI on your PATH produces a broken chat panel, and the README does not document how the plugin locates those binaries or what error you see when one is missing. Third, the README documents one performance workaround rather than a fix: on macOS with Retina displays, users are told to add -Dide.browser.jcef.gpu.disable=true to Help, Edit Custom VM Options and restart the IDE. Disabling JCEF GPU acceleration is a real trade-off, and the fact that it is the documented remedy suggests the embedded browser rendering path is the weak point. Finally, the README does not document rollback or downgrade for the plugin itself, so if a release regresses you should check CHANGELOG.md and the marketplace version list before upgrading.

## CC GUI versus running the CLI in a terminal or in VS Code

The direct alternative is the thing CC GUI wraps: run claude or codex in a terminal pane inside your IDE. That approach has no plugin to install, no webview, no Node bridge and no 40MB artifact, and it tracks the upstream CLI immediately rather than waiting for the plugin to be updated. What it lacks is the IDE integration CC GUI is selling: the code DIFF comparison view, file navigation and code jumping, session history with search and favorites, and message export. If you never use those, the plugin is overhead. The second alternative is a VS Code based GUI for the same CLIs. The difference is not features but the host: those extensions live in VS Code, and CC GUI is an IntelliJ IDEA plugin, so the choice is decided by which IDE you already use rather than by a feature comparison. The README does not claim parity with any VS Code extension, and the repository contains no VS Code extension code, so a VS Code user has no path to this plugin.

## Licence, release cadence and what maintenance costs you

The repository is MIT licensed, which permits commercial use, modification and redistribution provided the copyright notice and permission notice are retained. That is the standard MIT obligation, and it applies to the source in this repository; it says nothing about the licences of Claude Code, Codex or the other CLIs you install alongside it, which you must check separately. On maintenance, the last push was on 2026-09-20 and the most recent release listed is 0.5.6 on 2026-09-16, following 0.5.5 on 2026-09-01 and v0.5.4 on 2026-08-26. Releases are landing on a roughly two to three week cadence, and the version numbers are still in the 0.5.x range, which is a reasonable signal that the API surface is not yet frozen. The README also states that a /security-review audit runs before each minor version release and a comprehensive claude-code-security audit every 10 minor versions. That is a stated process, not a guarantee of outcomes, and it applies to the project's own code. Upgrade cost is the practical concern: because the plugin bundles a webview and a Node bridge and depends on external CLIs that move independently, a plugin update can break an engine that worked before. The README does not document a rollback procedure, so pinning a known-good marketplace version before upgrading is the safer habit.

## Conclusion

Adopt CC GUI if you already live in IntelliJ IDEA and want Claude Code or Codex sessions, diffs and history inside the IDE instead of a terminal tab. Skip it if you work primarily in VS Code, or if you need a stable interface for the Beta engines, since the README marks six of the eight engines as Beta and only Claude Code and Codex are listed without that label. Before committing, check the JetBrains Marketplace listing for the IDE versions the plugin declares as compatible, and confirm that the CLI each engine depends on is already installed and authenticated on your machine, because the plugin is a front end and does not ship those CLIs.

## FAQ

### Is JetBrains a Russian company?

The README does not cover JetBrains' corporate structure. CC GUI is a third-party plugin published on the JetBrains Marketplace and is not developed by JetBrains.

### Can I use the Claude Code GUI in IntelliJ?

Yes. CC GUI is an IntelliJ IDEA plugin that provides a visual interface for Claude Code and other AI coding CLIs, and the README points installation at its JetBrains Marketplace listing. You still need the underlying CLI available on your machine.

### Can JetBrains be trusted?

The README does not address that question. What it does state is that CC GUI runs a /security-review audit before each minor version release and a comprehensive claude-code-security audit every 10 minor versions.

### Is JetBrains better than VSCode?

The README does not compare the two IDEs. It does establish that CC GUI is an IntelliJ IDEA plugin with no VS Code counterpart in the repository, so the IDE choice decides whether this plugin is available to you at all.

## Sources

- [License: MIT](https://github.com/zhukunpenglinyutong/jetbrains-cc-gui/blob/main/LICENSE)
- [Project website](https://plugins.jetbrains.com/plugin/29342-cc-gui-claude-or-codex-)
- [README](https://github.com/zhukunpenglinyutong/jetbrains-cc-gui/blob/main/README.md)
- [Releases](https://github.com/zhukunpenglinyutong/jetbrains-cc-gui/releases)
- [zhukunpenglinyutong/jetbrains-cc-gui on GitHub](https://github.com/zhukunpenglinyutong/jetbrains-cc-gui)

---

Hysen Labs editorial analysis, written from the project's own repository and release notes. Cite the canonical page: https://hysenlabs.com/projects/zhukunpenglinyutong-jetbrains-cc-gui