中文文档 | Contributing | Documentation

Kode is a powerful AI assistant that lives in your terminal. It can understand your codebase, edit files, run commands, and handle entire workflows for you.

⚠️ Security Notice: Kode runs in YOLO mode by default (equivalent to --dangerously-skip-permissions), bypassing all permission checks. Use kode --safe to enable permission checks for important projects.

• 🤖 AI-Powered Assistance - Advanced AI models to understand and respond to your requests
• 🔄 Multi-Model Collaboration - Switch and combine multiple AI models (Option+M to cycle)
• 🦜 Expert Model Consultation - Use @ask-model-name for specialized analysis
• 👤 Intelligent Agent System - Use @run-agent-name to delegate tasks to subagents
• 📝 Code Editing - Direct file editing with intelligent suggestions
• 🔍 Codebase Understanding - Analyzes project structure and code relationships
• 🚀 Command Execution - Run shell commands in real-time
• 🛠️ Workflow Automation - Handle complex tasks with simple prompts
• 🎨 Interactive UI - Beautiful terminal interface with syntax highlighting
• 🔌 Extensible Tools - MCP servers, skills, and plugins

npm install -g @shareai-lab/kode

🇨🇳 China users: npm install -g @shareai-lab/kode --registry=https://registry.npmmirror.com

Dev channel: npm install -g @shareai-lab/kode@dev

Commands: kode (primary) | kwa (alternative) | kd (alias)

No WSL/Git Bash required. Kode downloads a native binary on postinstall and falls back to Node.js when needed.

Override: KODE_BINARY_BASE_URL (mirror) | KODE_SKIP_BINARY_DOWNLOAD=1 (skip)

# Interactive mode
kode

# Non-interactive mode
kode -p "explain this function" path/to/file.js

# ACP mode (for Toad/Zed)
kode-acp

• Global config: ~/.kode.json
• Project settings: ./.kode/settings.json
• Models: /model (UI) or kode models import/export (YAML)

Kode supports true multi-model collaboration - use different models for different tasks:

• Model Pointers: Configure main, task, compact, quick models via /model
• YAML Config: Export/import team-shareable configs with kode models export/import
• Expert Consultation: Use @ask-model-name to consult specific models
• Parallel Processing: Launch multiple subagents with @run-agent-name

# Export model config
kode models export --output kode-models.yaml

# Import
kode models import kode-models.yaml

Create agent templates in .kode/agents/:

---
name: reviewer
description: "Review diffs for correctness and security"
tools: ["Read", "Grep"]
model: inherit
---
Be strict. Point out bugs and risky changes.

Run: @run-agent-reviewer ... or Task(subagent_type: "reviewer", ...)

Install skills from any repository:

npx add-skill vercel-labs/agent-skills -a kode

Manage: /plugin marketplace add, /plugin install

Requires Bun:

# Install Bun
curl -fsSL https://bun.sh/install | bash  # macOS/Linux
powershell -c "irm bun.sh/install.ps1 | iex"  # Windows

# Setup
git clone https://github.com/shareAI-lab/kode.git
cd kode
bun install

# Development
bun run dev

# Build
bun run build

# Test
bun test

We welcome contributions! See Contributing Guide.

Apache 2.0 - see LICENSE.

• 📚 Documentation
• 🐛 Report Issues
• 💬 Discussions