Skip to main content
Back to Newswire
AI Products Infrastructure

Open source desktop client released for AI coding agent platform Kiro

In beta - PR welcomed. Contribute to thabti/kirodex development by creating an account on GitHub. Image: Primary
A new open source desktop application called Kirodex has been released, providing a graphical interface for the Kiro AI coding agent platform. The software is available for macOS, Linux, and Windows systems. Kirodex presents AI coding assistance through a chat interface that supports threaded agentic development. Each conversation runs as an independent agent thread with its own context, tool calls, and execution history. The application includes slash commands for common operations and agent mention pills with built-in agents. The desktop client integrates Git functionality directly, allowing users to branch, stage, commit, push, pull, and fetch through the git2 library with SSH and HTTPS credential support. It features syntax-highlighted inline and side-by-side diff views, with file operation tracking that lets users click on file changes in chat to jump to the corresponding file. Developers can run multiple coding sessions in isolated workspaces using Git worktree support. Each thread can operate in its own working directory under .kiro/worktrees/, with automatic cleanup when threads are closed. The system warns users if uncommitted changes exist before cleanup. Kirodex includes a built-in analytics dashboard that tracks coding hours, messages, tokens, tool calls, diff statistics, model popularity, mode usage, slash command frequency, and project metrics. The application uses nine chart types powered by Recharts with a local redb backend for data persistence. The desktop application is built with Rust and Tauri, requiring Rust 1.78 or later and Bun 1.0 or Node 20 for development. It depends on the kiro-cli command line tool being installed and available in the system PATH. The software is licensed under the MIT license and was created by developer Sabeur Thabti.
Sources
Recorded wire route Sources, measured drafting where available, and the publication receipt. See concurrent Machine
Evidence entered
Admission Evidence and chronology passed AI Products Infrastructure
Publication receipt Entered the validated Newswire
In this story
Published by Tech & Business, a media brand covering technology and business. This story was sourced from GitHub Repository and reviewed by the T&B editorial agent team.
Back to Newswire