Maintainer of this skill?
Automated security scan — not a certification.
by wxtsky
Real-time AI coding agent status panel in your MacBook notch — live status, approvals & replies for 30+ AI coding tools, with iPhone & Apple Watch companions
# Add to your Claude Code skills
git clone https://github.com/wxtsky/CodeIslandLast scanned: 7/19/2026
{
"issues": [],
"status": "PASSED",
"scannedAt": "2026-07-19T06:29:36.339Z",
"npmAuditRan": true,
"pipAuditRan": true,
"promptInjectionRan": true
}See how CodeIsland compares with popular alternatives.
CodeIsland is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by wxtsky. Real-time AI coding agent status panel in your MacBook notch — live status, approvals & replies for 30+ AI coding tools, with iPhone & Apple Watch companions. It has 2,502 GitHub stars.
Yes. CodeIsland passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.
Clone the repository with "git clone https://github.com/wxtsky/CodeIsland" and add it to your Claude Code skills directory (see the Installation section above).
CodeIsland is primarily written in Swift. It is open-source under wxtsky on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh CodeIsland against similar tools.
No comments yet. Be the first to share your thoughts!
30 days in the Featured rail · terms & refunds
⚠️ Third-Party Software Notice
This skill is third-party open-source software developed and hosted independently on GitHub. SkillsLLM is an informational directory and does not control or maintain the underlying repository.
Any security checks, ratings, or warnings displayed by SkillsLLM are automated and limited in scope. They do not constitute a security certification or guarantee that the software is safe, error-free, or free from malicious code, vulnerabilities, compromised dependencies, or prompt-injection risks.
Review the source code, permissions, dependencies, and configuration before installing or running any third-party skill. Use is at your own risk. To the maximum extent permitted by applicable law, SkillsLLM is not liable for losses arising from third-party software.
Automated security scan — not a certification.
Coding agents spend a lot of time either working or waiting on you — and you only find out by switching to their window. CodeIsland turns the notch into a live status bar for all of them: which session is thinking, which one needs an approval, which one just finished. Approve the tool call or answer the question right there, or click once to jump to the exact terminal tab.
It works with 30+ AI coding tools, installs its hooks for you, and keeps everything on your Mac.
👀 See everything at a glance
✋ Act without switching windows
🧘 Stays out of your way
🌐 Beyond this Mac
Also covered: Trae CN, Trae CLI / Trae CLI Next, Qoder CN, Qoder App, QoderWork and Qoder CLI, Cursor CLI, CodeBuddy CN, Claude Desktop (Code tab, and Cowork), ZCode — plus any tool with Claude-style hooks, added as a custom CLI in Settings → Hooks.
Several accounts? Register extra Claude Code, Codex and Grok config directories (e.g. a second CLAUDE_CONFIG_DIR) in Settings → Hooks; each gets its own hooks and status, and sessions, transcripts and usage cover all of them.
Knows where it runs: sessions inside tmux, zellij, Herdr or T3 Code get a chip next to the terminal badge, and click-to-jump goes to the right pane or thread.
Approvals & questions can be answered from the island for tools whose hooks wait for a decision — Claude Code, Codex, Gemini CLI, Qoder, Qwen Code, Trae CLI Next, ZCode, OpenCode, MiMo Code / Xiaomi MiMo, Pi / Oh My Pi, DeepSeek Harness and others; MiniMax Code CLI for the 10 seconds mcode lets a hook wait. Tools whose hooks can't carry a decision (Google Antigravity, AiWork) are shown read-only, and approvals stay in their own UI.
CodeIsland writes these for you on launch (and repairs them if they drift); each can be switched off in Settings → Hooks.
| Tool | Installed into |
|---|---|
| Claude Code | ~/.claude/settings.json (honours $CLAUDE_CONFIG_DIR) |
| Codex | ~/.codex/hooks.json — needs a one-time review |
| Gemini CLI | ~/.gemini/settings.json |
| Google Antigravity | ~/.gemini/config/hooks.json |
| Cursor / Cursor CLI | ~/.cursor/hooks.json |
| Grok CLI | ~/.grok/hooks/codeisland.json |
| Qoder (IDE and Qoder App) / Qoder CN / QoderWork | ~/.qoder/, ~/.qoder-cn/, ~/.qoderwork/ settings.json |
| Trae / Trae CN | ~/.trae/hooks.json, ~/.trae-cn/hooks.json (turn on global hooks in Trae) |
| Trae CLI |