Bunli

Build production-ready, type-safe CLIs with strong type inference, validation, and cross-platform distribution.

What is Bunli?

Bunli is a complete ecosystem for building command-line interfaces with Bun. It combines a minimal, type-safe framework with powerful tooling for development, testing, building, and distribution.

Type-Safe

Full TypeScript support with automatic type inference and code generation

// Types flow automatically
handler: async ({ flags }) => {
  // flags is fully typed
}

// Generated types for advanced patterns
import { getCommandApi } from './commands.gen'

Zero Dependencies

Pure implementation with no external dependencies

"dependencies": {}

Fast Startup

Leverages Bun's quick boot time for instant CLI response

$ time mycli --help
0.012s

Production Ready

Automated builds, releases, and cross-platform distribution

bunli build --all
bunli release

Quick Start

# Create a new CLI project
bunx create-bunli my-cli
cd my-cli

# Start development
bunli dev

# Build for production
bunli build

# Create a new CLI project
npx create-bunli my-cli
cd my-cli

# Start development
npm run dev

# Build for production
npm run build

Why Bunli?

Designed for Bun

Bunli is built specifically for the Bun runtime, taking advantage of:

Native TypeScript execution
Built-in test runner
Fast startup times
Bun Shell integration
Single-file executables

Type Inference First

Strong typing without manual annotations:

import { defineCommand, option } from '@bunli/core'
import { z } from 'zod'

// Bunli infers everything
export default defineCommand({
  name: 'deploy',
  options: {
    env: option(
      z.enum(['dev', 'staging', 'prod']),
      { description: 'Target environment' }
    )
  },
  handler: async ({ flags }) => {
    // TypeScript knows flags.env is 'dev' | 'staging' | 'prod'
    switch (flags.env) {
      case 'dev':     // ✅ Autocompleted
      case 'staging': // ✅ Autocompleted
      case 'prod':    // ✅ Autocompleted
    }
  }
})

Standard Schema Support

Use any validation library you prefer:

import { z } from 'zod'
import * as v from 'valibot'

// Use your favorite schema library
options: {
  email: option(z.string().email()),
  port: option(v.number([v.minValue(1), v.maxValue(65535)]))
}