codekanban

Code Kanban

中文 | English

Manage all your terminals and AI coding tools (Claude Code / Codex) from one unified page.

Core Features • Quick Start • Development Guide • About

Quick Start

Try it instantly:

npx codekanban
# or
npm install -g codekanban@latest
codekanban

Core Features

• 🚀 Ready to Use Single-file deployment, local database, double-click to run.

• 📊 Multi-Project & Multi-Terminal Management Manage all your projects and terminals from one unified page (Hotkey: `).

• 🔍 AI Coding Tool Status Detection Auto-detect running status of Claude Code / Codex and other tools (idle/working/needs approval), notify when tasks complete.

• 📜 History Record your conversations with AI and trace prompt history in one click.

• 🌿 Worktree Management A hybrid Git engine uses go-git for fast in-process status, diff, and line statistics, while preferring system Git for commits and repository mutations so hooks, signing, filters, and user configuration continue to work. Read and write engines can be selected independently in Settings.

• 📝 Notes Multi-tab notes for quick recording of ideas and issues (Hotkey: 1).

Installable Codex Skill Bundle

This repository ships an installable Codex skill bundle built around a single public CLI: codekanban-cli. Use it when you want the CodeKanban skill workflow on another machine without copying the full repository.

JavaScript package layout:

• packages/node-sdk: library-only SDK for JS integrations
• packages/codekanban-cli: the only public CLI and the owner of the packaged Codex skills
• packages/codekanban-cli/skills/codekanban-cli: the single shipped Codex skill source

Key defaults:

• Default service URL: http://127.0.0.1:3007
• First-time auth setup: printf '%s' '<PASSWORD>' | codekanban-cli auth save-token --password-stdin
• Non-default service URL: codekanban-cli --base-url http://192.168.1.50:3007 session list --path /repo
• Saved auth file: Windows %APPDATA%\codekanban-cli\session.json; macOS/Linux $XDG_CONFIG_HOME/codekanban-cli/session.json or ~/.config/codekanban-cli/session.json

The offline bundle installs one CLI package and copies one Codex skill. After installing the CLI and skills, restart Codex so the new skill is discovered.

Development Guide

Requirements

• Node.js: v20.19.0+ or v22.12.0+
• Go: 1.25+
• Package Manager: pnpm 9.15.9

Install Dependencies

Frontend Dependencies:

cd ui
pnpm install

Backend Dependencies:

go mod tidy

Development Mode

Frontend Development Server:

cd ui
pnpm dev

Access at: http://localhost:5173

Backend Development Server:

go run . # Note: After first run, config.yaml will be generated with port 3007. Due to conflicts with the production version, they cannot run simultaneously. It's recommended to change to 3005. The following assumes it has been changed.

• Server Port: http://localhost:3005
• OpenAPI Documentation: http://localhost:3005/docs
• Health Check: http://localhost:3005/api/v1/health

Optional Parameters:

• -m or --migrate: Force database migration
• -i or --install: Install as system service
• --uninstall: Uninstall system service

Production Build

Full Build (Recommended):

python build.py

This script will automatically complete the following steps:

1. Build frontend (pnpm build)
2. Copy frontend artifacts to static/ directory
3. Build Go executable (with optimization)

Manual Build:

# Build frontend
cd ui && pnpm build

# Build backend
go build -ldflags="-s -w" -trimpath -o CodeKanban

Build Artifacts:

• Frontend: ui/dist/ → static/ (After moving to this directory, building the backend will automatically store it in the executable, achieving single-file launch)
• Backend: CodeKanban.exe (Windows) or CodeKanban (Linux/macOS)

Access Application

Development Environment:

• Frontend Dev Server: http://localhost:5173
• Backend API: http://localhost:3005

Production Environment: Run the built executable and access http://localhost:3007

About

We are in an era that is both rapidly changing and extremely divided, and our work is the same.

This tool has genuinely improved my efficiency, but perhaps the improvement in efficiency will also reduce work demand, and the development of AI will eliminate this industry.

Anyway, I hope everyone enjoys using it.

If it helps you, you can give it a star or sponsor me a little.

Possible Future Features

• Mobile support
• Code cleanup: For example, the frontend's src/api should all be auto-generated
• Completion reminder feature, such as playing a sound after AI is done to tell you it's finished.
• Idle terminal list / Terminal list awaiting interaction.