Lumora

What it does

Lumora is a native macOS notch companion that brings AI coding-session status, music, and system information into a compact surface at the top of the screen. It is designed for developers who work across coding tools and want useful status without another full-size window.

The project includes integrations for Claude Code, Codex, OpenCode, and Cursor. Local agent events are organized into session timelines with prompts, tool activity, approvals, completion state, and token usage where supported. Music views provide now-playing metadata and playback controls. System views show CPU, memory, battery, and network status. English and Chinese interfaces are included, with music-driven colors, a black appearance, and Liquid Glass on supported systems.

How it works

Lumora is built with Swift and SwiftUI in Xcode. Provider-specific adapters process local hook events and transcript data, then feed a shared session-state pipeline for timeline and status views. Separate services handle music metadata, playback controls, and system sampling.

Technical challenges

Different coding providers expose different event behavior. Claude and Codex use transcript ordering, while OpenCode and Cursor can need temporary tool placeholders as events arrive. Completion, permission, and compaction states also require provider-specific handling. The public release history includes fixes for a localization-symbol collision and an ambiguous ProcessInfo reference.

Current release and testing

Lumora 1.0.0 was released on September 24, 2026, with a downloadable DMG. The recorded CI and release workflows passed. Automated tests cover provider adapters, transcript parsing, session-state transitions, media metadata mapping, and shared timeline updates.

The app requires macOS 15.6 or later. The release is ad-hoc signed and not notarized. Source code, setup instructions, and the download are linked below.

My contribution and AI usage

AI tools wrote the code. I completed all other aspects of the project, including defining the architecture and making technical decisions.

Submission preparation disclosure

AI assistance was used to review the public repository and draft this project description. This description summarizes repository-documented functionality; it is not a claim that every integration has been independently demonstrated in a live runtime session.

Demo scope and limitations

The accompanying demonstration was recorded on a real Mac with local build 1.4.0; the linked public release remains 1.0.0. The video discloses the version difference and an observed session-status badge that stayed Running after completion. Music playback is not demonstrated. Synthetic English narration and AI-assisted editing were used in preparing the demo.

Built With

• swift