CodeMoss: Multi‑AI Coding Companion for Desktop & IDE
CodeMoss: The Multi‑AI Companion Turning Code into Conversational Design
CodeMoss is a lightweight, cross‑platform desktop application that plugs directly into your development workflow. It aggregates several cloud‑hosted AI engines – Claude Code, Codex, Gemini, OpenCode and others – into one unified panel that feels native to VS Code, JetBrains IDEs, and the desktop itself.
Why a New AI Companion?
Traditional IDE extensions often lock you into a single AI model or a specific platform. CodeMoss flips that narrative by offering:
- Multi‑Model support – switch between Claude, Codex, Gemini, etc. on the fly.
- Cross‑IDE integration – plug into VS Code, JetBrains, or run standalone.
- Rich ecosystem – AI memory, superpowers, open‑spec APIs, spec‑kit for building your own plugins.
- No‑vendor lock‑in – all data runs locally unless you explicitly connect to a cloud model.
Key Features
| Feature | What it does | How it looks in CodeMoss |
|---|---|---|
| Chat Session | Interactive AI chat that can pull in your context and propose solutions. | 📺 chat window with command palette. |
| File Editor | Edit source files directly inside the panel, with syntax highlighting and version control hooks. | Sidebar file tree. |
| Git Integration | Visualise and manage Git, open commit diffs, and stage changes from the same UI. | Git panel shows branch status. |
| Terminal & CLI View | Render output from your terminal while the AI runs in the background. | Integrated terminal panel. |
| Superpowers | Auto‑complete prompts and offer context‑aware suggestions across your workspace. | AI assistant pops up contextually. |
| Parallel Task Execution | Run multiple AI tasks simultaneously (e.g., linting + code generation). | Multi‑tab panel. |
| Exportable AI History | Save chat logs or export to Markdown for documentation. | Export button. |
Getting Started
1. Download
- macOS, Windows, Linux: https://www.codemoss.ai/download
- Install using the standard installer or the
.dmg/.exe/.AppImagefiles.
2. Install the IDE Extension
- VS Code: https://marketplace.visualstudio.com/items?itemName=zhukunpenglinyutong.codemoss
- JetBrains: Install from the Marketplace or use the
codemossplugin.
3. Connect an AI provider
Open the settings panel and choose your provider (Claude, Gemini, Codex, or your own custom endpoint). You’ll be prompted for an API key and any additional configuration.
4. Start Chatting
Open the CodeMoss panel (Ctrl+Shift+M), type a prompt, hit Enter, and watch the AI generate responses live. You can reference files, ask for refactoring, or run custom scripts.
Architecture Overview
┌─────────────────────┐
│ Desktop Runtime │
├─────────────────────┤
│ Electron + Tauri │
│ (UI + Main Process)│
├─────────────────────┤
│ AI‑Bridge Layer │
│ (Rust + WASM APIs)│
├─────────────────────┤
│ External Providers │
│ (Claude, Codex, … )│
└─────────────────────┘
- Frontend – Built with Vue/Nuxt and TypeScript, leveraging the Electron renderer for platform‑native windows.
- Backend – Rust‑powered
src-tauricore exposes a clean API to the front‑end and abstracts provider communication. - Plugins – Spec‑kit lets others write modules in TypeScript or Rust, adding new AI endpoints or UI widgets.
Contributing
- All contributions welcome! Fork the repo, create a feature branch, and open a pull request.
- Tests are run via GitHub Actions – ensure you add CI scripts if you modify core logic.
- Documentation updates or new screenshots are always appreciated.
Roadmap
- Day‑by‑Day Iterations – At least one release per day for the next 30 days.
- Beta Version for Mobile – Experimental support for iOS/Android.
- More AI Models – Integration with emerging models (e.g., Llama‑2, GPT‑4o).
- Community Plug‑In Store – A curated list of community‑built extensions.
Community & Support
- Twitter: https://twitter.com/zhukunpenglinyutong
- Discord: https://discord.gg/CodeMoss (for real‑time help and feature requests)
- Issue Tracker: https://github.com/zhukunpenglinyutong/codemoss/issues
Feel free to star the repo, join the chat, and help shape the future of AI‑augmented coding. Each star fuels new features—let’s rewrite the workflow together!