Codeg (Code Generation) provides an enterprise-grade, multi-agent coding workspace designed to unify disparate local AI coding agents like Claude Code, Codex CLI, and Gemini CLI within a single desktop application. It targets engineers and power users seeking a streamlined environment for managing multiple AI coding tools, offering benefits such as session aggregation, parallel development workflows, and integrated Git/file/terminal operations.

How It Works

The project employs a hybrid architecture: a Next.js 16 (React 19) frontend with static export, communicating with a Tauri 2 backend written in Rust. This local-first design prioritizes privacy and security, with network access restricted to user-triggered actions and system proxy support for enterprise environments. The Rust backend manages agent communication (ACP), parses local sessions, handles Git/file/terminal runtimes, and facilitates an MCP marketplace and skills management, utilizing SeaORM with SQLite for local data persistence.

Quick Start & Requirements

• Primary Install/Run: Install dependencies with pnpm install, then run pnpm tauri dev for development or pnpm tauri build for a desktop build.
• Prerequisites: Node.js >=22 (recommended), pnpm >=10, Rust stable (2021 edition), and Tauri 2 build dependencies (e.g., libwebkit2gtk-4.1-dev, librsvg2-dev on Debian/Ubuntu Linux).
• Links: No specific quick-start or demo links are provided in the README.

Highlighted Details

• Unified workspace for local AI coding agents, supporting over 20 adapters including GitHub Copilot and Qwen Code.
• Session aggregation with structured rendering for historical AI coding sessions.
• Parallel development facilitated by built-in Git worktree flows.
• MCP (Multi-Agent Communication Protocol) and Skills management, supporting local scanning and registry installation.
• Integrated engineering loop encompassing file tree navigation, diff views, Git change tracking, commit operations, and terminal access.

Maintenance & Community

The project is currently at v0.1.x, indicating a phase of fast iteration suitable for early adopters. No specific community channels (e.g., Discord, Slack) or notable contributors/sponsorships are mentioned in the provided text.

Licensing & Compatibility

The project is licensed under the Apache-2.0 license. This permissive license generally allows for commercial use and integration within closed-source projects without significant restrictions.

Limitations & Caveats

As a v0.1.x release, the software is in rapid development and may be unstable. The frontend's static export limits Next.js dynamic routes, requiring the use of query parameters instead. Developers should note the differing parameter naming conventions between the frontend (camelCase) and the Rust backend (snake_case).